Task 3: Sequence-based predictions

From Bioinformatikpedia
Revision as of 22:59, 29 August 2011 by Meier (talk | contribs) (Results and discussion)

Contents

Task description

The full description of this task can be found here.


Task 3.1: Secondary structure prediction

PSIPRED

More information on PSIPRED can be found here: PSIPRED

Run with: sudo ./runpsipred reference.fasta

JPred3

More information on JPred can be found here: Jpred3

Used server: http://www.compbio.dundee.ac.uk/www-jpred/index.html

DSSP

More information on DSSP can be found here: DSSP

Run with: dssp 2PAH.pdb 2PAH.dssp

Result

It was necessary to align the sequence of the pdb-file and the fasta sequence of PAH.

Reference MSTAVLENPGLGRKLSDFGQETSYIEDNCNQNGAISLIFSLKEEVGALAK
1PHZ ------------------GQETSYIEDNSNQNGAISLIFSLKEEVGALAK
DSSP --------------------------------------------------
PSIPRED CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCEEEEEECCCCCHHHHH
JPred --HHHH--HHHHHHHHHH---------------EEEEEEEE----HHHHH
Reference VLRLFEENDVNLTHIESRPSRLKKDEYEFFTHLDKRSLPALTNIIKILRH
1PHZ VLRLFEENDINLTHIESRPSRLNKDEYEFFTYLDKRTKPVLGSIIKSLRN
DSSP --------------------------------------------------
PSIPRED HHHHHHHCCCCEEEEECCCCCCCCCCEEEEEECCCCCCHHHHHHHHHHCC
JPred HHHHHHH---EEEEEE----------EEEEEEEE---HHHHHHHHHHHHH
Reference DIGATVHELSRDKKKDTVPWFPRTIQELDRFANQILSYGAELDADHPGFK
1PHZ DIGATVHELSRDKEKNTVPWFPRTIQELDRFANQI------LDADHPGFK
DSSP -----------------.....SBGGGGGGTT.S.------..TTSTTTT
PSIPRED CCEEEECCCCCCCCCCCCCCCCCCCCHHHHHHHHHHHCCCCCCCCCCCCC
JPred H-----EEE----------------HHHHHH---EEE-------------
Reference DPVYRARRKQFADIAYNYRHGQPIPRVEYMEEEKKTWGTVFKTLKSLYKT
1PHZ DPVYRARRKQFADIAYNYRHGQPIPRVEYTEEEKQTWGTVFRTLKALYKT
DSSP .HHHHHHHHHHHHHHHH..TTS........HHHHHHHHHHHHHHHHHHHH
PSIPRED CHHHHHHHHHHHHHHHCCCCCCCCCCCCCCHHHHHHHHHHHHHHHHHHCC
JPred -HHHHHHHHHHHH-----------------HHHHHHHHHHHHHHHHH---
Reference HACYEYNHIFPLLEKYCGFHEDNIPQLEDVSQFLQTCTGFRLRPVAGLLS
1PHZ HACYEHNHIFPLLEKYCGFREDNIPQLEDVSQFLQTCTGFRLRPVAGLLS
DSSP HB.HHHHHHHHHHHHHS..BTTB...HHHHHHHHHHHT..EEEE.SS...
PSIPRED CHHHHHHHHHHHHHHHCCCCCCCCCCHHHHHHHHHHHHCCEEEECCCCCC
JPred --HHHHHHHHHHHHHH----------HHHHHHHHHHH---EEEE------
Reference SRDFLGGLAFRVFHCTQYIRHGSKPMYTPEPDICHELLGHVPLFSDRSFA
1PHZ SRDFLGGLAFRVFHCTQYIRHGSKPMYTPEPDICHELLGHVPLFSDRSFA
DSSP HHHHHHHHTTTEEEE......TT.TT..SS..HHHHHTTTTTTTTSHHHH
PSIPRED HHHHHHHCCCCEECCCEEEECCCCCCCCCCCCHHHHHHCCCCCCCCCHHH
JPred HHHHHHHH----EEEEEEE-----------HHHHHHHH--------HHHH
Reference QFSQEIGLASLGAPDEYIEKLATIYWFTVEFGLCKQGDSIKAYGAGLLSS
1PHZ QFSQEIGLASLGAPDEYIEKLATIYWFTVEFGLCKEGDSIKAYGAGLLSS
DSSP HHHHHHHHHHTT..HHHHHHHHHHHHTTTTT.EEEETTEEEE..HHHHT.
PSIPRED HHHHHHHHHCCCCCHHHHHHHHHHEEEEEEEEEECCCCCEEEECCCCCCC
JPred HHHHHHHHHHH---HHHHHHHHH-HHHEEEEEEEEE---EEEEE------
Reference FGELQYCLSEKPKLLPLELEKTAIQNYTVTEFQPLYYVAESFNDAKEKVR
1PHZ FGELQYCLSDKPKLLPLELEKTACQEYSVTEFQPLYYVAESFSDAKEKVR
DSSP HHHHHHTTSSSS..EE..HHHHTT....SSS..S..EEES.HHHHHHHHH
PSIPRED HHHHHHHHCCCCCCCCCCHHHHHCCCCCCCCCCEEEEEECCHHHHHHHHH
JPred HHHHHHHH-----EE---HHHHH-----------EEEE---HHHHHHHHH
Reference NFAATIPRPFSVRYDPYTQRIEVLDNTQQLKILADSINSEIGILCSALQK
1PHZ TFAATIPRPFSVRYDPYTQRVEVLDNT-----------------------
DSSP HHHHTS..SSEEEEETTTTEEEEE.HHHHHHHHHHHHHHHHHHHHHHHHH
PSIPRED HHHHHCCCCCEEEECCCCCEEEECCCHHHHHHHHHHHHHHHHHHHHHHHH
JPred HHHHHH------------EEEEE---HHHHHHHHHHHHHHHHHHHHHHHH
Reference IK
1PHZ --
DSSP T.
PSIPRED HC
JPred --

Discussion

DSSP can be regarded as a kind of reference for the secondary structure. It uses the coordinates of a resolved structure and is therefore much more reliable than PSIPRED or JPRED. There are more tools, which uses the resolved structure, but their predictions don't have to be the same. The results of these kind of methods are good hints for the secondary structure, but depend strongly on the used definitions of the different secondary structure elements. PSIPRED and JPRED predict the secondary structure only with the amino acid sequence of the protein. The advantage of these methods is that the resolved structure of the protein is not needed. To compare PSIPRED and JPRED with the DSSP result it is necessary to translate the 8-state prediction of DSST to a three state prediction. There are several possibilities to do this (see ref).

  • H (alpha-Helix), G (3_10 helix), I (pi-helix) -> H (alpha-helix)
  • E (extended strand) -> E (beta-strand)
  • B (residue in isolated beta-bridge), T (turn), S (bend), . (rest, coil) -> C (loop, coil)

There are two measures to evaluate secondary structure predictions: Q3 (true positives) and Segments OVerlapping (see ref). In this case the proteinmodel server was used to calculate the scores.

method - score ALL HELIX STRAND COIL
JPred - Q3 83.9 90.5 67.7 80.0
PSIPRED - Q3 83.9 89.9 71.0 80.0
JPred - SOV 84.9 91.8 66.6 81.0
PSIPRED - SOV 87.3 96.1 78.5 80.2

The difference between Jpred and PSIPRED is marginal. Both performed well on our sequence. Probably there was enough knowledge about our sequence in the training sets of these methods.

One of the major problems in secondary structure prediction are the misclassification of an observed helix as sheet and vise versa. The following tables show the frequency of the different classification and misclassification for the two prediction methods.

JPRED

observed/predicted frequency
HC 15
HH 143
CC 112
EC 10
EE 21
CE 16
CH 12

PSIPRED

observed/predicted frequency
HC 14
HH 142
CC 112
HE 2
EC 9
CE 14
EE 22
CH 14

JPred does not misclassify H to E or E to H. PSIPred misclassifies H to E only two times. Therefore the two prediction methods seem to be quite similar in quality.

Task 3.2: Prediction of disordered regions

DISOPRED

DISOPRED

Run with: ./rundisopred reference.fasta

POODLE

POODLE

Used server: http://mbs.cbrc.jp/poodle/poodle.html

IUPRED

IUPRED

Used server: http://iupred.enzim.hu/index.html

META-Disorder

META-Disorder

Used server: http://www.predictprotein.org/

Result

Position 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Sequence M S T A V L E N P G L G R K L S D F G Q E T
IUPRED long 0.3840 0.4051 0.4220 0.3356 0.3599 0.3872 0.2918 0.3149 0.3494 0.3807 0.4017 0.4256 0.3529 0.2715 0.3087 0.3740 0.4652 0.3910 0.3910 0.3704 0.3840 0.4652
IUPRED short 0.9447 0.8823 0.8457 0.8074 0.7540 0.6442 0.6035 0.5711 0.4458 0.4037 0.3668 0.4149 0.4116 0.4116 0.3578 0.3225 0.3939 0.3885 0.3885 0.2865 0.2913 0.3630
MD D D D - - - - - - - - - - - - - - - - - - -
DisoPred D D D D D D D D D D D D D D D D D D D D D D
Poodle D D D - - - - D D D D D D D D D D D D D D D
Position 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
Sequence S Y I E D N C N Q N G A I S L I F S L K E E
IUPRED long 0.4441 0.3704 0.3704 0.3704 0.2988 0.1969 0.2884 0.2129 0.1969 0.2002 0.2064 0.1554 0.2258 0.2783 0.1969 0.1731 0.1554 0.1554 0.0967 0.0888 0.0518 0.0269
IUPRED short 0.3630 0.3668 0.3456 0.3359 0.3359 0.2385 0.2385 0.1602 0.1532 0.0832 0.0771 0.0858 0.1416 0.1532 0.1416 0.1117 0.0643 0.0813 0.0789 0.0502 0.0308 0.0297
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred D D D D D D D D D D - - - - - - - - - - - -
Poodle D D D D D D D - - - - - - - - - - - - - - -
Position 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
Sequence V G A L A K V L R L F E E N D V N L T H I E
IUPRED long 0.0334 0.0592 0.0662 0.1070 0.1184 0.2094 0.1476 0.2193 0.2258 0.1476 0.1449 0.1942 0.1914 0.2436 0.3117 0.3321 0.2645 0.3182 0.3910 0.3983 0.4864 0.5139
IUPRED short 0.0218 0.0218 0.0231 0.0414 0.0723 0.1266 0.0789 0.1380 0.0935 0.0832 0.0771 0.0677 0.0701 0.0858 0.1322 0.1844 0.1878 0.2385 0.2255 0.2292 0.3053 0.3939
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
Sequence S R P S R L K K D E Y E F F T H L D K R S L
IUPRED long 0.5098 0.3948 0.2817 0.2817 0.3460 0.2575 0.3426 0.3321 0.3249 0.4017 0.3182 0.3494 0.3286 0.2258 0.2328 0.2503 0.2503 0.1823 0.1823 0.1184 0.0719 0.1137
IUPRED short 0.3885 0.3053 0.3096 0.2209 0.2167 0.2122 0.3005 0.2167 0.2122 0.3005 0.2865 0.2913 0.2167 0.2209 0.1998 0.1495 0.2209 0.2292 0.1635 0.1088 0.1088 0.1088
MD - D D - - - - - - - - - - - - - - - - - - -
DisoPred - - D D D D D - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110
Sequence P A L T N I I K I L R H D I G A T V H E L S
IUPRED long 0.1070 0.1852 0.2064 0.2034 0.1759 0.2503 0.1881 0.1942 0.2034 0.1424 0.2034 0.1759 0.2002 0.2680 0.2575 0.2364 0.3149 0.3948 0.3392 0.4409 0.4051 0.3019
IUPRED short 0.0643 0.1060 0.1602 0.1566 0.0991 0.1667 0.1495 0.1456 0.1060 0.1041 0.1635 0.0832 0.0965 0.1416 0.1456 0.1456 0.2041 0.2786 0.2820 0.3535 0.3535 0.4037
MD - - - - - - - - - - - - - - - - - - - - - D
DisoPred - - - - - - - - - - - - - - - - D D D D D D
Poodle - - - - - - - - - - - - - - - - - - - - - D
Position 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132
Sequence R D K K K D T V P W F P R T I Q E L D R F A
IUPRED long 0.3215 0.3149 0.3948 0.3149 0.3494 0.3566 0.3392 0.3426 0.3286 0.2988 0.2752 0.2918 0.2988 0.2224 0.1583 0.1611 0.1007 0.1007 0.1229 0.1137 0.1349 0.2292
IUPRED short 0.3146 0.2385 0.3578 0.3399 0.3578 0.2913 0.3762 0.3885 0.3184 0.4116 0.4037 0.3263 0.3053 0.3096 0.3359 0.2748 0.2122 0.2167 0.2167 0.1844 0.2483 0.3399
MD D D D - - - - - - - - - - - - - - - - - - -
DisoPred D D D D D D D D - - - - - - - - - - - - - -
Poodle D D - - - - - - - - - - - - - - - - - - - -
Position 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
Sequence N Q I L S Y G A E L D A D H P G F K D P V Y
IUPRED long 0.1823 0.1881 0.1852 0.2849 0.2752 0.1643 0.2292 0.2292 0.2575 0.3053 0.2503 0.2364 0.2002 0.2752 0.3494 0.3494 0.4409 0.3321 0.3286 0.3215 0.3182 0.2918
IUPRED short 0.2167 0.2333 0.2041 0.2820 0.2657 0.2963 0.3630 0.2700 0.2786 0.3668 0.4245 0.3578 0.2602 0.3184 0.3491 0.3456 0.4116 0.3992 0.4458 0.3668 0.4333 0.4333
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
Sequence R A R R K Q F A D I A Y N Y R H G Q P I P R
IUPRED long 0.2328 0.2436 0.1671 0.1449 0.1399 0.2328 0.2470 0.2680 0.1702 0.2470 0.3249 0.2645 0.3019 0.2193 0.1852 0.2002 0.1969 0.3149 0.3356 0.3286 0.4119 0.3286
IUPRED short 0.4078 0.3847 0.3096 0.2748 0.1998 0.2786 0.2786 0.2657 0.2432 0.3311 0.3263 0.3535 0.3885 0.3263 0.3263 0.2558 0.1958 0.2602 0.2820 0.2748 0.3311 0.3399
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198
Sequence V E Y M E E E K K T W G T V F K T L K S L Y
IUPRED long 0.4119 0.4017 0.4256 0.3215 0.3215 0.3249 0.2364 0.2094 0.2849 0.1852 0.1206 0.1643 0.1583 0.2292 0.2399 0.1476 0.0851 0.0851 0.0506 0.0568 0.0531 0.0662
IUPRED short 0.4037 0.2786 0.3491 0.3491 0.2865 0.1998 0.1878 0.1635 0.1635 0.1349 0.1322 0.1240 0.0771 0.1150 0.1349 0.1240 0.1060 0.0542 0.0316 0.0363 0.0212 0.0464
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220
Sequence K T H A C Y E Y N H I F P L L E K Y C G F H
IUPRED long 0.0334 0.0380 0.0341 0.0398 0.0443 0.0424 0.0405 0.0235 0.0244 0.0198 0.0313 0.0356 0.0364 0.0405 0.0287 0.0605 0.1092 0.0618 0.1137 0.1115 0.0765 0.1184
IUPRED short 0.0425 0.0268 0.0182 0.0414 0.0245 0.0128 0.0259 0.0259 0.0137 0.0087 0.0160 0.0084 0.0081 0.0094 0.0128 0.0279 0.0279 0.0308 0.0554 0.0308 0.0376 0.0660
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242
Sequence E D N I P Q L E D V S Q F L Q T C T G F R L
IUPRED long 0.2129 0.1048 0.1028 0.1671 0.1583 0.0929 0.1554 0.2399 0.1373 0.2328 0.1643 0.1528 0.1643 0.1070 0.1424 0.1229 0.0704 0.0704 0.0618 0.0568 0.0851 0.0929
IUPRED short 0.0771 0.0621 0.1150 0.1088 0.0554 0.0643 0.1060 0.1266 0.1292 0.2080 0.1380 0.1240 0.0660 0.0643 0.1349 0.0701 0.0526 0.0991 0.0490 0.0268 0.0441 0.0441
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264
Sequence R P V A G L L S S R D F L G G L A F R V F H
IUPRED long 0.0424 0.0518 0.0851 0.0799 0.0454 0.0690 0.0320 0.0320 0.0424 0.0218 0.0334 0.0184 0.0165 0.0244 0.0178 0.0117 0.0174 0.0258 0.0269 0.0263 0.0253 0.0275
IUPRED short 0.0414 0.0789 0.0701 0.0395 0.0395 0.0744 0.0464 0.0376 0.0405 0.0425 0.0514 0.0297 0.0200 0.0173 0.0252 0.0274 0.0286 0.0304 0.0304 0.0268 0.0259 0.0259
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286
Sequence C T Q Y I R H G S K P M Y T P E P D I C H E
IUPRED long 0.0443 0.0424 0.0433 0.0506 0.0948 0.1323 0.2399 0.1611 0.1323 0.2193 0.2436 0.2645 0.1791 0.1702 0.2470 0.2609 0.2849 0.2094 0.1184 0.1184 0.1323 0.1137
IUPRED short 0.0502 0.1041 0.0813 0.0464 0.0771 0.1117 0.1958 0.2209 0.2963 0.3184 0.2255 0.3184 0.3311 0.2333 0.2483 0.3311 0.3263 0.2786 0.3096 0.2385 0.1667 0.1416
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308
Sequence L L G H V P L F S D R S F A Q F S Q E I G L
IUPRED long 0.1349 0.1184 0.0948 0.0851 0.0364 0.0300 0.0300 0.0592 0.0662 0.0690 0.0356 0.0494 0.0817 0.0483 0.0518 0.0646 0.0662 0.1184 0.2292 0.1554 0.0909 0.1028
IUPRED short 0.1958 0.2748 0.2786 0.1698 0.1322 0.1240 0.0660 0.1041 0.1878 0.2167 0.1240 0.1844 0.1635 0.1349 0.1349 0.1088 0.0909 0.1088 0.1958 0.2041 0.2041 0.1322
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
Sequence A S L G A P D E Y I E K L A T I Y W F T V E
IUPRED long 0.1007 0.1229 0.1229 0.1229 0.1449 0.0929 0.0454 0.0205 0.0349 0.0235 0.0433 0.0263 0.0275 0.0275 0.0165 0.0191 0.0174 0.0165 0.0160 0.0248 0.0244 0.0214
IUPRED short 0.0789 0.1456 0.1602 0.0884 0.1178 0.1117 0.0567 0.0286 0.0327 0.0160 0.0327 0.0252 0.0297 0.0304 0.0286 0.0194 0.0078 0.0066 0.0055 0.0090 0.0194 0.0157
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352
Sequence F G L C K Q G D S I K A Y G A G L L S S F G
IUPRED long 0.0188 0.0184 0.0218 0.0178 0.0281 0.0287 0.0327 0.0646 0.0605 0.0463 0.0405 0.0929 0.0555 0.0967 0.1007 0.0592 0.0327 0.0320 0.0327 0.0327 0.0662 0.0662
IUPRED short 0.0086 0.0157 0.0157 0.0083 0.0157 0.0226 0.0350 0.0376 0.0200 0.0350 0.0304 0.0308 0.0316 0.0621 0.0677 0.0771 0.0455 0.0268 0.0128 0.0131 0.0308 0.0316
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
Sequence E L Q Y C L S E K P K L L P L E L E K T A I
IUPRED long 0.0483 0.0494 0.0581 0.0424 0.0473 0.0483 0.0929 0.0909 0.0967 0.1501 0.0929 0.0870 0.1501 0.0888 0.1501 0.1671 0.2470 0.2715 0.1501 0.1611 0.1611 0.1048
IUPRED short 0.0526 0.0909 0.0514 0.0395 0.0490 0.0490 0.0991 0.0554 0.0643 0.1088 0.1088 0.0813 0.0813 0.0813 0.1349 0.1495 0.2167 0.1532 0.1322 0.1495 0.0701 0.0789
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396
Sequence Q N Y T V T E F Q P L Y Y V A E S F N D A K
IUPRED long 0.1048 0.1028 0.0568 0.0765 0.0765 0.1229 0.0543 0.0618 0.0662 0.0646 0.1048 0.1115 0.0985 0.1070 0.1028 0.1611 0.0870 0.0631 0.1007 0.1007 0.0483 0.0985
IUPRED short 0.1566 0.1698 0.0965 0.1060 0.0567 0.0858 0.0789 0.0832 0.0502 0.0643 0.1205 0.0965 0.0909 0.1456 0.1416 0.1456 0.1456 0.1416 0.1495 0.1205 0.0858 0.1292
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418
Sequence E K V R N F A A T I P R P F S V R Y D P Y T
IUPRED long 0.1554 0.2575 0.2164 0.2470 0.1671 0.1702 0.1969 0.1969 0.2064 0.1643 0.1759 0.1611 0.1671 0.1528 0.1671 0.1092 0.1251 0.1583 0.2002 0.2002 0.2951 0.2715
IUPRED short 0.1205 0.2041 0.2786 0.3456 0.2432 0.2483 0.2602 0.2558 0.1766 0.2122 0.2558 0.1805 0.1844 0.2657 0.2657 0.1732 0.2657 0.2209 0.1667 0.1805 0.2657 0.2292
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
Sequence Q R I E V L D N T Q Q L K I L A D S I N S E
IUPRED long 0.1942 0.1643 0.1852 0.1275 0.1554 0.1702 0.2399 0.1501 0.1424 0.2094 0.2193 0.1399 0.1399 0.1424 0.0835 0.0799 0.1275 0.0985 0.0555 0.0568 0.0518 0.0294
IUPRED short 0.2385 0.2122 0.2080 0.2041 0.1878 0.1240 0.1416 0.1456 0.1240 0.1041 0.1041 0.1018 0.1088 0.0991 0.0965 0.0832 0.0813 0.0514 0.0526 0.0425 0.0200 0.0182
MD - - - - - - - - - - - - - - - - - - - - - -
DisoPred - - - - - - - - - - - - - - - - - - - - - -
Poodle - - - - - - - - - - - - - - - - - - - - - -
Position 441 442 443 444 445 446 447 448 449 450 451 452
Sequence I G I L C S A L Q K I K
IUPRED long 0.0463 0.0473 0.0662 0.0870 0.0888 0.0618 0.0483 0.0765 0.0494 0.0364 0.0214 0.0356
IUPRED short 0.0363 0.0173 0.0316 0.0744 0.1240 0.1456 0.1766 0.3146 0.3578 0.3992 0.4333 0.5802
MD - - - - - - - - D D D D
DisoPred - - - - - - - - - - - -
Poodle - - - - - - - - - - D D

Discussion

It's difficult to compare the different methods, because there is no reference. In the database Disprot (Disprot) you can find the disordered regions of different proteins. In a database search with our reference sequence we found one reliable match with DP00094. If this alignment is right, there should be a disordered region in our protein from residue 30 to 107. IUPRED measures the disorder with a kind of score between 0 and 1. It's hard to define a threshold. Assuming that IUPRED predicts a disorder in this regions the threshold is about 0.25. Of course this can only be an approximation.

method beginning region central region end region
MD 1-3 110-113 449-452
DisoPred 1-32 105-118 none
Poodle 1-29 110-112 451-552
IUPRED long 1-29 136-182 none
IUPRED short 1-27 136-181 448-452

The methods seem to hit something in the neighborhood of the the region, which is disordered according to the Disprot. The data on disorder and therefore the methods' prediction trained on this data is not very accurate. Perhaps with more accurate experiments to identify disorder, prediction methods get a better base.

According to the hints in the oral discussion of this task, we compare the predictions to the secondary structure and the b-factors of the CA. Most experimental structures known for PAH are limited to the residues 110 to 424. For this comparison we used the experimental structure 1J8U for PAH.

Figure 1: This figure shows the residues of the structure 1j8u of PAH colored by the b-factor of their C-alpha atom.

Figure 1 shows the residues of PAH colored by the b-factor of their C-alpha atom. The b-factor is a measure of the flexibility of an atom in an experimental structure. There are three regions, which seem to be of higher flexibility.

region secondary structure remark
118-120 coil begin of the experimental structure
130-135 alpha-helix
422-424 beta-sheet end of the experimental structure

We were surprised by this visualization. In fact we did not expect to see the greatest flexibility in structured regions like an alpha-helix or a beta-sheet. The flexibility of the peripheral and solvent exposed parts of the structures can be somehow explained. But the flexibility of the alpha-helix residues 130 to 135 are interesting. In nature PAH is a tetra-homo-mere. It consists of a regulatory N-terminal domain (residues 1-117), the catalytic domain (residues 118-427), and a C-terminal domain (residues 428-453) responsible for oligomerization of the identical monomers. The flexible region 422 to 424 is probably more defined in the tetra-homo-mere form. PAH seems to increase in volume during its catalytic activity. PAH is usually a homo-tetra-mere connected by the C-terminal domain. Therefore it seems reasonable, that the linker residues 422 to 424 are more flexible. The linker is probably not disordered, but the connecting domain can be.

There were four other regions with high flexibility flanking the binding pocket within the catalytic domain.

region secondary structure
247-248 coil
378-381 coil
145 coil
135 sheet

These regions are probably flexible due to the catalytic function of the domain. They are probably more defined during certain steps of the catalytic activity of the protein. Their flexibility is probably necessary to keep the solvent away from the binding pocket. At least no method predicted these sites to be disordered.

Figure 2: This figure shows the residues of the structure 1phz of PAH colored by the b-factor of their C-alpha atom.

A more complete experimental structure, 1PHZ, of the rat PAH is shown in figure 2. The structure shows the N-terminal regulatory domain. This domain seems to be very unstructured and flexible. The first 20 amino acids could not be located in the experiment, which indicates a very high flexibility. This domain is usually excluded in experiments, because it is proposed to be too flexible and increases the susceptibility to proteases. Therefore the prediction of disorder in this region is probably right.








In the following we pay respect to the secondary structure of the regions predicted to be disordered.

method region secondary structure secondary structure origin
MD 1-3 CC JPred
MD 110-113 CCCC JPred
MD 449-452 HHCC DSSP
DisoPred 1-32 CCHHHHCCHHHHHHHHHHCCCCCCCCCCCCCC JPred
DisoPred 105-118 CCEEECCCCCCCCC JPred
Poodle 1-29 CCHHHHCCHHHHHHHHHHCCCCCCCCCCC JPred
Poodle 110-112 CCC JPred
Poodle 451-552 CC DSSP
IUPRED long 1-29 CCHHHHCCHHHHHHHHHHCCCCCCCCCCC JPred
IUPRED long 136-182 ------CCCCCCCCCCHHHHHHHHHHHHHHHHCCCCCCCCCCCCCH DSSP
IUPRED short 1-27 CCHHHHCCHHHHHHHHHHCCCCCCCCC JPred
IUPRED short 136-181 ------CCCCCCCCCCHHHHHHHHHHHHHHHHCCCCCCCCCCCCCH DSSP
IUPRED short 448-452 HHHCC DSSP

Every method hit parts of the protein with defined structure (helix, sheet). The secondary structure was evaluated by DSSP, where it was possible. DSSP uses an experimental structure. In our case it was 1PHZ. 1PHZ contains a Fe-atom. Perhaps this is enough to introduce order in disordered regions of the protein. As seen in the experimental structure 1PHZ, the N-terminal domain is unstructured. Therefore the secondary structure prediction of JPred in this region is probably wrong.

In order to summarize the results:

  • There is probably disorder in the N-terminal regulatory domain. The first 20 amino acids of 1PHZ were not able to be detected in the experiment and this domain is usually excluded in experimental structures, because it is too flexible and attracts proteases. This indicates disorder in the N-terminal region. This was predicted by the methods. The experimental structure of 1PHZ shows that this region is mostly unstructured and of high flexibility.
  • There is probably no disorder in the catalytic domain. Some of the surface loops are flexible, but this flexibility is probably part of the catalytic process.
  • The C-terminal domain, which is necessary to build the homo-tetra-mere, may contain disorder, but it is unlikely. The domain seems to be structurally defined. The flexible linker residues 422 to 424 are flexible due to the expansion of the catalytic domain during the procession of phenylalanine.

The prediction of disorder seems to be in the fledgling stages.

Task 3.3: Prediction of transmembrane alpha-helices and signal peptides

Annotated sequence features

PAH

The phenylalanine-4-hydroxylase has no annotated signal peptide or transmembrane helices.

BACR_HALSA

The bacteriorhodopsin has the following annotated signal peptide and transmembrane helices:

Position Feature Name Description
1 - 13 Propeptide
14 – 23 Topological domain Extracellular
24 - 42 Transmembrane Helical; Name=Helix A
43 – 56 Topological domain Cytoplasmic
57 - 75 Transmembrane Helical; Name=Helix B
76 – 91 Topological domain Extracellular
92 - 109 Transmembrane Helical; Name=Helix C
110 – 120 Topological domain Cytoplasmic
121 - 140 Transmembrane Helical; Name=Helix D
141 – 147 Topological domain Extracellular
148 - 167 Transmembrane Helical; Name=Helix E
168 – 185 Topological domain Cytoplasmic
186 - 204 Transmembrane Helical; Name=Helix F
205 – 216 Topological domain Extracellular
217 - 236 Transmembrane Helical; Name=Helix G
237 – 262 Topological domain Cytoplasmic


RET4_HUMAN

The retinol-binding protein 4 has the following annotated signal peptide (no transmembrane helices are annotated):

Position Feature Name Description
1 - 18 Signal peptide

INSL5_HUMAN

The Insulin-like peptide INSL5 has the following annotated signal peptide (no transmembrane helices are annotated):

Position Feature Name Description
1 - 22 Signal peptide


LAMP1_HUMAN

The lysosome-associated membrane glycoprotein 1 has the following annotated signal peptide and transmembrane helices:

Position Feature Name Description
1 - 28 Signal peptide
29 – 382 Topological domain Lumenal
383 - 405 Transmembrane Helical;
406 – 417 Topological domain Cytoplasmic


A4_HUMAN

The Amyloid beta A4 protein has the following annotated signal peptide and transmembrane helices:

Position Feature Name Description
1 - 17 Signal peptide
18 – 699 Topological domain Extracellular
700 - 723 Transmembrane Helical;
724 – 770 Topological domain Cytoplasmic

General Questions to prediction of transmembrane alpha-helices and signal peptides

Why is the prediction of transmembrane helices and signal peptides grouped together here?

Methods which only predict transmembrane helices often predict signal peptides as transmembrane helices as well. The reason for this is that both, transmembrane helices and signal peptides consist mainly of hydrophobic residues. These false predictions lead to inaccurate topological features and thus to wrongly annotated function of a protein. To avoid these cases most recent methods couple their transmembrane prediction together with a signal peptide prediction.

Description of different signal peptides

Signalpeptides for the import to the endoplasmic reticulum (ER)

The import to the ER is usually required for the secretory pathway (to export proteins out of a cell). The import process can occur either co-translational (the nascent protein chain is translocated together with the ribosome) or post-translational (only the fully synthesized protein is transported to the ER). However, for both cases the SEC-pathway is mostly used.

The co-translational transport to the ER is done by the signal recognition particle (SRP). This particle recognizes the N-terminal signal-sequence of the nascent polypeptide chain and then transports it to the ER membrane where the complex, consisting of SRP, polypeptide chain and ribosome, is recognized by the ER membrane bound signal recognition particle receptor (SR). After this recognition the polypeptide chain is imported into the ER lumen via the SEC channel in an ATP dependent process.

The post-translational import to the ER lumen is done by chaperons which guide the polypeptide chain to the SEC channel which is then imported in an ATP dependent process.

However, not only the import to the ER lumen is possible, an import to the ER membrane is possible as well. So far, 5 different types of import to the ER membrane are known.

Type 1 requires an N-terminal signal sequence and an intrinsic stop transfer anchor sequence which will be the part which is inserted in the membrane.

Type 2 and 3 do not require a N-terminal signal sequence only a intrinsic signal anchor sequence is required. The difference between type 2 and 3 is that type 2 has positively charged residues before the signal anchor sequence (on the N-Terminal side) and type 3 has positively charged residues after the signal anchor sequence (on C-Terminal side). These charged residues of trans-membrane protein are always in the cytosol. Thus, type 2 inserted proteins have their N-terminal end residing in the cytosol whereas type 3 inserted proteins have a C-terminal end in the cytosol.

Type 4-A and 4-B insertion is also known as multipass membrane insertion. These proteins have not one trans-membrane helix like the proteins imported via Type 1,2 and 3, instead they have several trans-membrane helices. Hence, they consist of multiple internal stop-transfer anchor sequences and internal signal-anchor sequences. The difference between type 4-A and 4-B is that in type 4-A the N and C terminal ends are located in the cytosol whereas type 4-B import results in a N-terminal end residing in the ER lumen and a C-terminal end residing in the cytosol.

In addition to the N-terminal import of trans-membrane proteins there is also the possiblity for a C-terminal import. Obviously, these proteins are imported post-translation.

Signalpeptides for the import to the mitochondrion

There are several targets for import to the mitochondrion, proteins can be translocated to the matrix, the outer membrane, the inner membrane and the inter membrane space.

Proteins who are designated to be imported to the matrix of a mitochondrion have a N-terminal matrix-targeting sequence. This mitochondrial import to the matrix is assisted by chaperons (Hsc70) which guide the protein to the import pore complex of the mitochondrion. The import through the outer membrane is conducted by the TOM complex and the following import through the inner membrane is conducted by the TOM complex. After successful import to the matrix the signal sequence is cleaved off by proteolytic active enzymes.

Import to the inner membrane can occur in three ways. The first way is the TIM22 pathway, proteins using this pathway need internal targeting sequences. The next way is the stop transfer import, for this proteins need a stop transfer sequence and a N-terminal matrix targeting sequence. The third way is called conservative sorting proteins using this pathway have a N-terminal targeting sequence as well and in addition intrinsic Oxa1-targeting sequences which are recognized by Ox1-proteins which execute the import to the membrane.

Proteins imported to the outer membrane of a mitochondrion usually have PORTA domains which are recognized by the TOB/SAM complex.

Signalpeptides for the import to the chloroplast

Proteins heading to chloroplasts can target different parts of it. For example the stroma, inner and outer membrane, the thylakoids membrane or the thylakoids lumen.

Usually these protein have a N-terminal targeting sequence.

Signalpeptides for the import to the peroxisome

Peroxisomal proteins can be imported to the lumen or to the membrane. Proteins imported to the lumen have either a peroxisomal targeting signal at the C-termins (also known as PTS1) or a targeting sequence close to the N-terminus (also known as PTS2). Proteins imported to the membrane can have an intrinsic membrane peroxisomal targeting signal (mPTS). However, not all proteins have this mPTS. These proteins are imported to the ER and from there they bud off together with the mature peroxisome.

Signalpeptides for the import to the nucleus and the export form the nucleus

Proteins which are imported to the nucleus require a nuclear localisation signal (NLS) which is recognized by importin. The NLS containing protein is then imported via the nuclear pore complex (NPC) to the nucleoplasm.

Proteins which are exported from the nucleus require a nuclear export signal which is recognized by exportin, a protein which binds to the NES of the cargo protein. In addition to exportin a second component, known as Ran*GTP, is required to mediate the export through the NPC.

TMHMM

Details of the method

Author: Sonnhammer, Heijne & Krogh

Year: 1998

Reference: PubMed

Description
Figure 3: HMM architecture of TMHMM Disclaimer: This file is redistributed from [Krogh A, Larsson B, von Heijne G, Sonnhammer EL. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol. 2001 Jan 19;305(3):567-80.] . All rights belong to the creator.

This method is based on a hidden markov model (HMM). The authors of this method tried to model the 'grammar' of transmembrane proteins in order to predict the protein topology of transmembrane more accurate than methods who only e.g. rely on propensity values and do not consider the topological constraints of these class of proteins.

TMHMM defined for their HMM for each feature one or more states which present this feature. For example the transmembrane helix is modeled by three sub models. A model for the helix core, the cap of the helix which lies partly in the cytoplasm and the membrane and the cap which is partly in the membrane and cytoplasm. In addition to this helix model they also created sub models for the cytoplasmic loop and the non-cytoplasmic loop as well as a sub model for the globular region. Each sub model can reflect one or more states in the HMM model. For example the globular sub model only consists of one HMM state whereas the helix-core and caps are modeled by multiple HMM states.

The 'grammar' is incorporated to this HMM model by defining the possible transitions from one sub model to another one. For example it is only possible to change from a cytoplasmic loop region to a cytoplasmic cap region and then to the helix core and after that either to non-cytoplasmic short loop or long non-cytoplasmic loop and so on.

Predicted features

This methods predicts the transmembrane helix and whether this part is in the cytoplasm (in) or outside of it (out).

Required information for the prediction

User who want to use it just need their amino acid sequence of their query sequence. The transmission and emission probabilities are derived from 160 transmembrane protein sequences.


Execution

Before we could execute TMHMM we had to change all occurrences of "/usr/local/bin/" to "/usr/bin" in these files: tmhmm, tmhmm.ORIG and tmhmmformat.pl

Then we executed the following command to retrieve the results for all sequences:

  • tmhmm all.fa > task_33/tmhmm_out.txt

Results and discussion

PAH
Position Feature Name
1 - 452 outside

TMHMM predicted no transmembrane helix as expected. However, what TMHMM predicted is that this protein is outside (outside of cytosol). Which is wrong if we look at the annotations from UniProt which says that it appears in the cytosol.


BACR_HALSA
Position Feature Name
1 - 22 outside
23 - 42 TMhelix
43 - 54 inside
55 - 77 TMhelix
78 - 91 outside
92 - 114 TMhelix
115 - 120 inside
121 - 143 TMhelix
144 - 147 outside
148 - 170 TMhelix
171 - 189 inside
190 - 212 TMhelix
213 - 262 outside

The transmembrane helices of BACR_HALSA were almost correctly predicted. The predicted positions of the transmembrane helices differ only by few positions. However, TMHMM failed to predict the last helix (Helix G, from 217 - 236) and hence also the C-terminal end was falsly predicted to be outside which is actually inside the cytoplasm.

RET4_HUMAN
Position Feature Name
1 - 201 outside

TMHMM predicted no transmembrane helix which is according to UniProt correct. Also the predicted cellular location (outside) is correctly predicted.


INSL5_HUMAN
Position Feature Name
1 - 135 outside

TMHMM predicted no transmembrane helix which is according to UniProt correct. Also the predicted cellular location (outside) is correctly predicted.

LAMP1_HUMAN
Position Feature Name
1 - 10 inside
11 - 33 TMhelix
34 - 383 outside
384 - 406 TMhelix
407 - 417 inside

TMHMM predicted the first transmembrane helix wrongly (position 11 - 33) which is not existent with reference to Uniport. However, the second predicted transmembrane helix is correct. As an effect of the wrongly predicted first transmembrane helix the N-terminal and was predicted to be inside (cytosol) which is actually outside (lumenal).

A4_HUMAN
Position Feature Name
1 - 700 outside
701 - 723 TMhelix
724 - 770 inside

TMHMM predicted all transmembrane helices and the topology of this protein correctly.

Phobius

Details of the method

Author: Käll, Krogh, Sonnhammer

Year: 2004

Reference: PubMed

Description
Figure 4: HMM architecture of Phobius Disclaimer: This file is redistributed from [KKäll L, Krogh A, Sonnhammer EL. A combined transmembrane topology and signal peptide prediction method. J Mol Biol. 2004 May 14;338(5):1027-36.] . All rights belong to the creator.

Phobius is an HMM based prediction method to predict transmembrane helices as well as N-terminal signal peptides. More precisely, it is a combination of the two HMM models of TMHMM and SignalP which is merged into one HMM. This was done in order to overcome problems associated with transmembrane helix prediction: signale peptides are often wrongly predicted as transmembrane helices. The complete architecture can be seen in the figure.

Predicted features

Phobius predicts transmembrane helices, signal peptides and the topology of the loops (whether they are inside the cytoplasm or not).

Required information for the prediction

Users only has to enter the amino acid sequence of their query protein in FASTA format.

Execution

We used the standard parameter of Phobius and submitted the sequences of all requested proteins as one fasta file. As seen in figure 4.

Figure 5: A screenshot of the input form of Phobius.

Results and discussion

Figure 6: A screenshot of the output of Phobious.
PAH

Phobius predicted the protein to be non cytoplasmic which is wrong. The location of PAH is in the cytoplasm as stated by UniProt.

BACR_HALSA

Phobius predicted all transmembrane helices as well as the topology of BACR_HALSA correctly. Only the boundaries differed slightly to the annotated ones in UniProt.

RET4_HUMAN

The signal peptide as well as the topology was predicted correctly by Phobius.

INSL5_HUMAN

The signal peptide as well as the topology was predicted correctly by Phobius.

LAMP1_HUMAN

The signal peptide, transmembrane helices and the topology was predicted correctly by Phobius.

A4_HUMAN

The signal peptide, transmembrane helices and the topology was predicted correctly by Phobius.

PolyPhobius

Details of the method

Author: Käll L, Krogh A, Sonnhammer EL

Year: 2005

Reference: PubMed

Description

PolyPhobius is also based on a HMM which constraints the possible transitions from one state to another in order to reflect the 'grammar' of transmembrane proteins. However, the difference to the ordinary Phobius is that it uses knowledge homologous sequences of the query sequences as well to make the prediction more accurate.

In order to do so it calculates for each sequence position for each label (e.g. transmembrane helix, in, out, etc...) for each homologous sequence the posterior label probability (PLP). The PLP is defined as "the probability of a label at a certain position in the sequence, given the sequence and the model" (quoted from "Käll L, Krogh A, Sonnhammer EL. An HMM posterior decoder for sequence feature prediction that includes homology information Bioinformatics. 2005 Jun;21 Suppl 1:i251-7."). Then a multiple sequence alignment (MSA) of all homologous sequences is build, for each position in the MSA a average PLP is calculated. This average PLP will be then be used by the optimal accuracy algorithm to predict the most likely sequences of states for a given query sequence and thus the topology of the transmembrane helices.

Predicted features

This method predicts the same features as the ordinary Phobius, which means transmembrane helices, the signal peptide and whether the connecting loops of transmembrane helices are inside or outside.

Required information for the prediction

User need the amino acid sequence of their protein in FASTA format. An additional option is to specify the homologous sequences manually. If that is not done PolyPhobius will search for homologous sequences by itself by using BLAST.

Execution

We used the standard parameter of PolyPhobius and submitted the sequences of all requested proteins as one fasta file. As seen in the following picture:

Polyphobius all in.png

Results and discussion

Polyphobius all out.png


PAH

PolyPhobius predicted the protein to be non cytoplasmic which is wrong. The location of PAH is in the cytoplasm as stated by UniProt.

BACR_HALSA

PolyPhobius predicted all transmembrane helices as well as the topology of BACR_HALSA correctly. Only the boundaries differed slightly to the annotated ones in UniProt.

RET4_HUMAN

The signal peptide as well as the topology was predicted correctly by PolyPhobius.

INSL5_HUMAN

The signal peptide as well as the topology was predicted correctly by PolyPhobius.

LAMP1_HUMAN

The signal peptide, transmembrane helices and the topology was predicted correctly by PolyPhobius.

A4_HUMAN

The signal peptide, transmembrane helices and the topology was predicted correctly by PolyPhobius.

OCTOPUS

Details of the method

Author: Viklund H, Elofsson A.

Year: 2008

Reference: Bioinformatics

Description
Flowchart of OCTOPUS Disclaimer: This file is redistributed from [Viklund H, Elofsson A. OCTOPUS: improving topology prediction by two-track ANN-based preference scores and an extended topological grammar. Bioinformatics. 2008 Aug 1;24(15):1662-8. Epub 2008 May 12.] . All rights belong to the creator.

OCTOPUS basically uses two methods to predict the topology of transmembrane proteins: artificial neural networks (ANN) and hidden markov models (HMM). In a first step BLAST searches for homologous sequences of a input FASTA sequence. From the found homologous sequences a multiple sequence alignment is build from which a raw sequence profile and a sequence profile based on PSSM are extracted. These profiles are used for two sets of ANNs.

The first set of ANNs contains four separate ANNs which predict the residue preference for M (Membrane), I (Interface), L (Loop), G (Globular). In order to make the predictions for G and M more smooth the output of the first row of ANNs output is used for a second ANN as input.The second set of ANNs is taken to predict the residue preferences for the inside/outside residues.

Finally. the output of these two sets of ANNs are used to parameterize the OCTOPUS-HMM for the actual topological feature prediction. This HMM is needed to model the 'grammar' of trans membrane proteins, which simply means that only certain state transitions are allowed. For example, if we assume we are currently in the transmembrane state then it is only allowed to go into the loop state and so on and so forth.

The state sequences which fits best the input sequence is then calculated by the Viterbi algorithm.

Predicted features

Predicted features are inside/outside (i/o), transmembrane (M), TM hairpin (H), reentrant (R) or membrane dip (D)

Required information for the prediction

Only the amino acid sequence of the users protein is required.

Execution

We entered for each protein the amino acid sequence in fasta format. The picture below shows the procedure for PAH:

Octopus pah in.png

Results and discussion

PAH

Octopus pah out.png

No transmembrane helix was predicted. This is correct accordingly to the UniProt annotations.

BACR_HALSA

Octopus bacr halsa out.png

Octopus predicted all transmembrane helices as well as the topology of BACR_HALSA correctly.

RET4_HUMAN

Octopus RET4 HUMAN out.png

Octopus predicted the signal peptide to be a transmembrane helix which is incorrect. However, the topology (outside) was predicted correctly.

INSL5_HUMAN

Octopus INSL5 human out.png

Octopus predicted the signal peptide to be a transmembrane helix which is incorrect. However, the topology (outside) was predicted correctly.

LAMP1_HUMAN

Octopus LAMP1 HUMAN out.png

Octopus predicted an transmembrane helix close to the N-terminal end which is not correct. Only the second predicted transmembrane helic is correct as well as the topology.

A4_HUMAN

Octopus a4 human out.png

Octopus predicted a reentran/dip region close to the N-terminal end which is not annotated by UniProt. However, the rest is correct (second predicted transmemrane helix and overall topology).

SPOCTOPUS

Details of the method

Author: Viklund H, Bernsel A, Skwark M, Elofsson A.

Year: 2008

Reference: Bioinformatics

Description

SPOCTOPUS works the same way as OCTOPUS does. The only difference is that it includes a signal peptide prediction.

Predicted features

Predicted features are signal peptide, inside/outside (i/o), transmembrane (M), TM hairpin (H), reentrant (R) or membrane dip (D)


Required information for the prediction

Only the amino acid sequence of the query protein is required as input.

Execution

We entered for each protein the amino acid sequence in fasta format. The picture below shows the procedure for PAH:

Octopus pah in.png

Results and discussion

PAH

Spoctopus pah out.png

No transmembrane helix was predicted. This is correct accordingly to the UniProt annotations.

BACR_HALSA

Spoctupus bacr halsa out.png

Spoctopus predicted all transmembrane helices as well as the topology of BACR_HALSA correctly.

RET4_HUMAN

Spoctopus RET4 HUMAN out.png

Spoctopus predicted the signal peptide as well as the topology correctly.

INSL5_HUMAN

Spoctopus INSL5 human out.png

Spoctopus predicted the signal peptide as well as the topology correctly.

LAMP1_HUMAN

Spoctopus LAMP1 HUMAN out.png

Spoctopus predicted the signal peptide, the transmembrane helix as well as the topology correctly.

A4_HUMAN

Spoctopus a4 human out.png

Spoctopus predicted the signal peptide, the transmembrane helix as well as the topology correctly.

SignalP

Details of the method

Author: Henrik Nielsen, Jacob Engelbrecht, Søren Brunak and Gunnar von Heijne.

Year: 1997

Reference: PubMed

Description

This predictor takes two methods into account the first method used is a neural network the second is a hidden markov model.

There are two neural networks one which is predicting whether the first n amino acids belong to a signal peptide and the second network predicts the exact cleavage side positon.

In a later version of SignalP a hidden markov model (HMM) has been also build to predict signal peptides. However, this prediction is completely independent from the neural network prediction. This HMM models the N-terminal region of a signal peptide as well as the surrounding cleavage site.

Predicted features

Predicts the presence of signal peptidase I cleavage sites and whether the first n residues belong to a signal peptide.

Required information for the prediction

The amino acid sequence of the protein and whether this protein is from a eukaryote, gram-negative bacteria or gram-positive bacteria.

Execution

Before we could execute SignalP on our virtual machine we had to change the path of the signalp file to /apps/signalp-3.0

Then we executed for each protein the following commands:

  • signalp -format short -t euk PAH.fa > task_33/signalp_pah_out
  • signalp -format short -t euk A4_HUMAN.fa > task_33/signalp_a4_human_out
  • signalp -format short -t gram- BACR_HALSA.fa > task_33/signalp_bacr_halsa_out
  • signalp -format short -t euk LAMP1_HUMAN.fa > task_33/signalp_lamp1_human_out
  • signalp -format short -t euk RET4_HUMAN.fa > task_33/signalp_ret4_human_out
  • signalp -format short -t euk INSL5_HUMAN.fa > task_33/signalp_insl5_human_out

Results and discussion

PAH

SignalP predicted in both methods (HMM and NN) that there is no cleavage site for an signal peptide. Which is correct with respect to the annotations in UniProt.

BACR_HALSA

SignalP predicted in both methods (HMM and NN) that there is no cleavage site for an signal peptide. Which is correct with respect to the annotations in UniProt.


RET4_HUMAN

SignalP predicted in both methods (HMM and NN) the cleavage site at positions 19. Which is correct with respect to the annotations in UniProt.

INSL5_HUMAN

SignalP predicted in both methods (HMM and NN) the cleavage site at positions 23. Which is correct with respect to the annotations in UniProt.

LAMP1_HUMAN

SignalP predicted in both methods (HMM and NN) the cleavage site at positions 29. Which is correct with respect to the annotations in UniProt.

A4_HUMAN

SignalP predicted in both methods (HMM and NN) the cleavage site at positions 18. Which is correct with respect to the annotations in UniProt.

TargetP

Details of the method

Author: Henrik Nielsen, Jacob Engelbrecht, Søren Brunak and Gunnar von Heijne.

Year: 1997

Reference: PubMed

Description
Architecture of TargetP Disclaimer: This file is redistributed from Emanuelsson O, Nielsen H, Brunak S, von Heijne G. Predicting subcellular localization of proteins based on their N-terminal amino acid sequence. J Mol Biol. 2000 Jul 21;300(4):1005-16.] . All rights belong to the creator.

TargetP's prediction are based on trained neural networks. These neural networks are build up in a two layer setup. The first layer consists of three neural networks which are used to predict whether it is a chloroplast targeting sequence, a mitochondrial targeting sequence or a signal peptide. The output of this first layer is then used in the second layer neural network as input to make the final prediction. Then the decision unit decides whether the cutoffs are obeyed. The output is then one of three classes cTP/mTP/SP/other and a reliability class value (RC) which is an indicator for the predictions certainty.

However, if a non-plant protein is entered the prediction for cTP is not applied for obvious reasons.

Predicted features

Predicts the localization to the following targets: chloroplast, mitochondrion, ER/golgi/secreted, and "other".

Required information for the prediction

The amino acid sequence of the protein and whether this protein is from a plant or non-plant organism.

Execution

We used the standard parameter of TargetP and submitted all sequences via one fasta file to the prediction server.

Targetp in.png

Results and discussion

Targetp out.png

PAH

TargetP predicted no target protein. This is correct with respect to the annotations from UniProt.

BACR_HALSA

TargetP predicted a signal peptide of length 116 and that the protein will be secreted. This is incorrect with respect to the annotations from Uniprot. These say that there is no target protein and that this protein is integrated in the cytoplasmic membrane.

RET4_HUMAN

TargetP predicted a signal peptide of length 18 and that the protein will be secreted. This is correct with respect to the annotations from UniProt.

INSL5_HUMAN

TargetP predicted a signal peptide of length 22 and that the protein will be secreted. This is correct with respect to the annotations from UniProt.

LAMP1_HUMAN

TargetP predicted a signal peptide of length 28 and that the protein will be secreted. This is correct with respect to the annotations from UniProt.

A4_HUMAN

TargetP predicted a signal peptide of length 17 and that the protein will be secreted. This is correct with respect to the annotations from UniProt.

Task 3.4: Prediction of GO terms

Annotated sequence features

PAH

The phenylalanine-4-hydroxylase has the following annotated GO terms:

Class GO Identifier GO Name
Function GO:0003824 catalytic activity
Function GO:0004497 monooxygenase activity
Function GO:0004505 phenylalanine 4-monooxygenase activity
Function GO:0005506 iron ion binding
Component GO:0005829 cytosol
Process GO:0006558 L-phenylalanine metabolic process
Process GO:0006559 L-phenylalanine catabolic process
Process GO:0006571 tyrosine biosynthetic process
Process GO:0008152 metabolic process
Process GO:0008652 cellular amino acid biosynthetic process
Process GO:0009072 aromatic amino acid family metabolic process
Function GO:0016491 oxidoreductase activity
Function GO:0016597 amino acid binding
Function GO:0016714 oxidoreductase activity, acting on paired donors, with incorporation or reduction of molecular oxygen, reduced pteridine as one donor, and incorporation of one atom of oxygen
Process GO:0018126 protein hydroxylation
Process GO:0034641 cellular nitrogen compound metabolic process
Process GO:0042136 neurotransmitter biosynthetic process
Process GO:0042423 catecholamine biosynthetic process
Process GO:0042558 pteridine-containing compound metabolic process
Function GO:0042803 protein homodimerization activity
Process GO:0046146 tetrahydrobiopterin metabolic process
Function GO:0046872 metal ion binding
Function GO:0048037 cofactor binding
Process GO:0055114 oxidation-reduction process

BACR_HALSA

The bacteriorhodopsin has the following annotated GO terms:

Class GO Identifier GO Name
Function GO:0004872 receptor activity
Function GO:0005216 ion channel activity
Component GO:0005886 plasma membrane
Process GO:0006810 transport
Process GO:0006811 ion transport
Process GO:0007602 phototransduction
Function GO:0009881 photoreceptor activity
Process GO:0015992 proton transport
Component GO:0016020 membrane
Component GO:0016021 integral to membrane
Process GO:0018298 protein-chromophore linkage
Process GO:0050896 response to stimulus

RET4_HUMAN

The retinol-binding protein 4 has the following annotated GO terms:

Class GO Identifier GO Name
Process GO:0001654 eye development
Function GO:0005215 transporter activity
Function GO:0005488 binding
Function GO:0005501 retinoid binding
Function GO:0005515 protein binding
Component GO:0005576 extracellular region
Component GO:0005615 extracellular space
Process GO:0006094 gluconeogenesis
Process GO:0006810 transport
Process GO:0007283 spermatogenesis
Process GO:0007507 heart development
Process GO:0007601 visual perception
Process GO:0008584 male gonad development
Process GO:0009790 embryo development
Function GO:0016918 retinal binding
Function GO:0019841 retinol binding
Process GO:0030277 maintenance of gastrointestinal epithelium
Process GO:0030324 lung development
Process GO:0032024 positive regulation of insulin secretion
Process GO:0032526 response to retinoic acid
Process GO:0032868 response to insulin stimulus
Function GO:0034632 retinol transporter activity
Process GO:0034633 retinol transport
Process GO:0042572 retinol metabolic process
Process GO:0042574 retinal metabolic process
Process GO:0042593 glucose homeostasis
Process GO:0045471 response to ethanol
Process GO:0048562 embryonic organ morphogenesis
Process GO:0048706 embryonic skeletal system development
Process GO:0048738 cardiac muscle tissue development
Process GO:0048807 female genitalia morphogenesis
Process GO:0050896 response to stimulus
Process GO:0050908 detection of light stimulus involved in visual perception
Process GO:0051024 positive regulation of immunoglobulin secretion
Process GO:0060041 retina development in camera-type eye
Process GO:0060044 negative regulation of cardiac muscle cell proliferation
Process GO:0060059 embryonic retina morphogenesis in camera-type eye
Process GO:0060065 uterus development
Process GO:0060068 vagina development
Process GO:0060157 urinary bladder development
Process GO:0060347 heart trabecula formation

INSL5_HUMAN

The insulin-like peptide INSL5 has the following annotated GO terms:

Class GO Identifier GO Name
Function GO:0005179 hormone activity
Component GO:0005575 cellular_component
Component GO:0005576 extracellular region
Process GO:0008150 biological_process

LAMP1_HUMAN

The lysosome-associated membrane glycoprotein 1 has the following annotated GO terms:

Class GO Identifier GO Name
Component GO:0005624 membrane fraction
Component GO:0005764 lysosome
Component GO:0005765 lysosomal membrane
Component GO:0005768 endosome
Component GO:0005770 late endosome
Component GO:0005771 multivesicular body
Component GO:0005886 plasma membrane
Component GO:0005887 integral to plasma membrane
Process GO:0006914 autophagy
Component GO:0009897 external side of plasma membrane
Component GO:0009986 cell surface
Component GO:0010008 endosome membrane
Component GO:0016020 membrane
Component GO:0016021 integral to membrane
Component GO:0031982 vesicle
Component GO:0042383 sarcolemma
Component GO:0042470 melanosome

A4_HUMAN

The amyloid beta A4 protein has the following annotated GO terms:

Class GO Identifier GO Name
Process GO:0000085 G2 phase of mitotic cell cycle
Process GO:0001967 suckling behavior
Process GO:0002576 platelet degranulation
Function GO:0003677 DNA binding
Function GO:0004867 serine-type endopeptidase inhibitor activity
Function GO:0005102 receptor binding
Function GO:0005488 binding
Function GO:0005515 protein binding
Component GO:0005576 extracellular region
Component GO:0005624 membrane fraction
Component GO:0005737 cytoplasm
Component GO:0005794 Golgi apparatus
Component GO:0005886 plasma membrane
Component GO:0005887 integral to plasma membrane
Component GO:0005905 coated pit
Process GO:0006378 mRNA polyadenylation
Process GO:0006417 regulation of translation
Process GO:0006468 protein phosphorylation
Process GO:0006878 cellular copper ion homeostasis
Process GO:0006897 endocytosis
Process GO:0006915 apoptosis
Process GO:0006917 induction of apoptosis
Process GO:0007155 cell adhesion
Process GO:0007176 regulation of epidermal growth factor receptor activity
Process GO:0007219 Notch signaling pathway
Process GO:0007409 axonogenesis
Process GO:0007596 blood coagulation
Process GO:0007617 mating behavior
Process GO:0007626 locomotory behavior
Process GO:0008088 axon cargo transport
Function GO:0008201 heparin binding
Process GO:0008219 cell death
Process GO:0008344 adult locomotory behavior
Process GO:0008542 visual learning
Component GO:0009986 cell surface
Process GO:0010466 negative regulation of peptidase activity
Process GO:0010952 positive regulation of peptidase activity
Component GO:0016020 membrane
Component GO:0016021 integral to membrane
Process GO:0016199 axon midline choice point recognition
Process GO:0016322 neuron remodeling
Process GO:0016358 dendrite development
Function GO:0016504 peptidase activator activity
Component GO:0019717 synaptosome
Process GO:0030168 platelet activation
Process GO:0030198 extracellular matrix organization
Function GO:0030414 peptidase inhibitor activity
Component GO:0030424 axon
Process GO:0030900 forebrain development
Component GO:0031093 platelet alpha granule lumen
Process GO:0031175 neuron projection development
Component GO:0031410 cytoplasmic vesicle
Component GO:0031594 neuromuscular junction
Function GO:0033130 acetylcholine receptor binding
Process GO:0035235 ionotropic glutamate receptor signaling pathway
Component GO:0035253 ciliary rootlet
Process GO:0040014 regulation of multicellular organism growth
Function GO:0042802 identical protein binding
Component GO:0043005 neuron projection
Component GO:0043197 dendritic spine
Component GO:0043198 dendritic shaft
Component GO:0043231 intracellular membrane-bounded organelle
Process GO:0045087 innate immune response
Component GO:0045177 apical part of cell
Component GO:0045202 synapse
Process GO:0045665 negative regulation of neuron differentiation
Process GO:0045931 positive regulation of mitotic cell cycle
Process GO:0045944 positive regulation of transcription from RNA polymerase II promoter
Function GO:0046872 metal ion binding
Component GO:0048471 perinuclear region of cytoplasm
Process GO:0048669 collateral sprouting in absence of injury
Process GO:0050803 regulation of synapse structure and activity
Process GO:0050885 neuromuscular process controlling balance
Process GO:0051124 synaptic growth at neuromuscular junction
Component GO:0051233 spindle midzone
Process GO:0051402 neuron apoptosis
Function GO:0051425 PTB domain binding
Process GO:0051563 smooth endoplasmic reticulum calcium ion homeostasis

GOPET

Details of the method

Author: Vinayagam A, König R, Moormann J, Schubert F, Eils R, Glatting KH, Suhai S

Year: 2004

Reference: PubMed

Description
Flowchart of GOPET Disclaimer: This file is redistributed from [Vinayagam A, del Val C, Schubert F, Eils R, Glatting KH, Suhai S, König R. GOPET: a tool for automated predictions of Gene Ontology terms. BMC Bioinformatics. 2006 Mar 20;7:161.] . All rights belong to the creator.

The prediction of GO terms is based on support vector machine (SVM) predictions. The training of this SVM was done with 39,740 selected GO-annotated cDNA sequences. For each of this training sequence they extract all annotated GO terms. In a next step they search for homologous sequences with blast with a e-value < 0.01. Sequences which fulfill this condition are used to extract attributes: including sequence similarity meas- ures, such as e-value, bitscore, identity, coverage score, alignment length, GO-term frequency, GO-term relationships between homologues, the level of annotation within the GO hierarchy and annotation quality of the homologues.

These attributes are then assigned to each GO term found in the training sequence. The training of the SVM is then done by taking the GO term and its associated attributes to train the SVM.

After the training the SVM is capable to predict GO terms from unknown cDNA or protein sequences in the same fashion.


Predicted features

GOPET predicts the GO term together with a confidence value.

Required information for the prediction

The cDNA or amino acid sequence of the protein is required.

Execution

We used the standard parameter of GOPET and submitted all requested protein sequences at once to the server.

Gopet all in.png

Results and discussion

PAH

Gopet pah out.png

Gopet predicted falsely the following GO terms: GO:0004510, GO:0004511, GO:0008199, GO:0008198. These GO terms are not annotated by UniProt.

BACR_HALSA

Gopet bacr halsa out.png

Gopet predicted falsely the following GO terms: GO:0008020, GO:0015078. These GO terms are not annotated by UniProt.

RET4_HUMAN

Gopet ret4 human out.png

Gopet predicted falsely the following GO terms: GO:0008289, GO:0005319 and GO:0008035. These GO terms are not annotated by UniProt.

INSL5_HUMAN

Gopet insl5 human out.png

Gopet predicted no false GO terms.

LAMP1_HUMAN

Gopet lamp1 human out.png

Gopet predicted falsely the following GO terms: GO:0004812, GO:0005524. These GO terms are not annotated by UniProt.

A4_HUMAN

Gopet a4 human out.png

Gopet predicted falsely the following GO terms: GO:003568, GO:0030304, GO:0030414, GO:0008270, GO:0005507 and GO:0005506. These GO terms are not annotated by UniProt.

Pfam

Details of the method

Author: Wellcome Trust Sanger Institute and Howard Hughes Janelia Farm Research Campus

Year: latest release in March 2011

Reference: Oxford Journals

Description

Pfam is a protein family sequence database. In order to build families a seed sequence alignment of homologous sequences is build which all belong to the same family. This alignment is then used to build a profile hidden markov model (HMM) which is then represent one family. These profile HMM can then be used to search in your query sequence or in sequence database for significant family matches. The tool used to do all this is HMMER3.

Predicted features

Pfam predicts protein families.

Required information for the prediction

The amino acid sequence of the protein.

Execution

We used the standard parameter for Pfam and submitted all sequences in one fasta file to the prediction server.

Pfam all in.png

Results and discussion

Pfam all out.png


PAH

Pfam could predict the domain ACT from 36 to 88. This domain is also annotated in UniProt. Although the position of this domain is not correct, the correct position is from 35 to 110. Secondly, Pfam predicted another domain wich is called Biopterin_H which is not annotated by UniProt.

BACR_HALSA

Pfam could predict the domain Bacteriorhodopsin from 23 to 253. This domain is also annotated in UniProt. Although the position of this domain is not correct, the correct position is from 14 to 262. Secondly, Pfam predicted another domain wich is called DUF21 which is not annotated by UniProt.

RET4_HUMAN

Pfam predicted the domains DspF and Lipocalin from 12 to 60 and from 39 to 173 respectively. Although these domains are not annotated by UniProt. In addition Pfam was not able to predict the domains Retinol-binding protein 4 and Plasma retinol-binding protein which are annotated by UniProt.

INSL5_HUMAN

Pfam predicted the domain Insulin from 27 to 135. This seems to be a correct prediction since INSL5_HUMAN is a insulin protein. However the correct annotated domains by UniProt are Insulin-like peptide INSL5 B chain and Insulin-like peptide INSL5 A chain.

LAMP1_HUMAN

Pfam predicted the domain Lamp from 29 to 109 and from 111 to 417 as two domains. In reality these domains are annotated as one by UniProt. In addition Pfam predicted the domain DUF1180 which is not annotated by UniProt.

A4_HUMAN

Pfam predicted the domains APP_N, Beta-App and APP_amyloid. For these domains there seems to be a annotation of it in UniProt. Although the predicted position for these features differ highly from the annotated ones. In additon Pfam predicted the following domains which are not annotated in UniProt: APP_Cu_bd, Kunitz_BPTI, APP_E2, Exonuc_VII_L and Activator_TraM.

ProtFun 2.2

Details of the method

Author: L. Juhl Jensen, R. Gupta, N. Blom, D. Devos, J. Tamames, C. Kesmir, H. Nielsen, H. H. Stærfeldt, K. Rapacki, C. Workman, C. A. F. Andersen, S. Knudsen, A. Krogh, A. Valencia and S. Brunak.

Year: 2002

Reference: PubMed

Description

The prediction of GO terms is based on a neural network. The training sequence set was obtained from looking for protein families and their assigned GO terms in the InterPro database and then mapping these InterPro domain matches to SWISS-PROT and TrEMBL to get the actual sequence information. In order to avoid over-fitting a homology reduction was performed afterwards. Then a set of 16 features for each sequence was derived which include features such as propeptide cleavage site predictions and subcellular compartment predictions from TargetP.

Then the training to the neural network was applied to find out the best weight for each feature and GO term. However, after extensive training they figured out that the method gives only reliable predictions to 14 GO categories and thus only these were selected to be predicted by the neural network.


Predicted features

ProtFun predicts the cellular role, whether the protein is a enzyme or not, the enzyme class and the Gene ontology category. The predicted gene ontology categories are :

  • Signal transducer
  • Receptor
  • Hormone
  • Structural protein
  • Transporter
  • Ion channel
  • Voltage-gated ion channel
  • Cation channel
  • Transcription
  • Transcription regulation
  • Stress response
  • Immune response
  • Growth factor
  • Metal ion transport


Required information for the prediction

Only the amino acid sequence of the protein is required.

Execution

We used the standard parameter and submitted all requested sequences at once.

Protfun all in.png

Results and discussion

PAH

Protfun pah out.png

ProtFun was not able to predict one of the 14 gene ontology categories with significance (no arrow). Although, ProtFun predicted the protein to be involved in amino acid biosynthesis and to be an enzyme which is correct. However, it false predicted PAH to be an isomerase.

BACR_HALSA

Protfun bacr halsa out.png

Protfun predicted BACR_HALSA to be involved in transport and binding which is correct. The second prediction that it is no enzyme is also correct. Laslty, the third prediction, ion channel is also correct. Thus, all three significant predictions made by ProtFun are correct.

RET4_HUMAN

Protfun ret4 human out.png

ProtFun predicted as function to be a cell envelop for which I did not find any evidence in UniProt. Second it said it is a enzyme which is also not correct since it is a transporter protein. Third it predicted this protein belongs to the enzyme class lyase for which I also found no evidence. However the gene ontology category immune response seems to be predicted correctly.

INSL5_HUMAN

Protfun insl5 human out.png

ProtFun predicted as function to be a cell envelop for which I did not find any evidence in UniProt. The prediction noenzyme seems to be correct for me. The gene ontology category prediction was hormone which is also correct.

LAMP1_HUMAN

Protfun lamp1 human out.png

ProtFun predicted as function to be a cell envelop for which I did not find any evidence in UniProt. The prediction noenzyme seems to be correct for me. The gene ontology category prediction was immune response which is also correct.

A4_HUMAN

Protfun a4 human out.png

The functional category prediction was cell envelope for which I did not find any evidence. Secondly it predicted this protein to be a enzyme which is not correct. UniProt declared this protein to be a receptor. Thirdly, it says it is a structural protein which is also not correct.