Difference between revisions of "Task 8 (MSUD)"

From Bioinformatikpedia
m (Visualization of mutations)
m (Summary)
 
(9 intermediate revisions by 2 users not shown)
Line 6: Line 6:
 
<gallery widths=450px heights=250px perrow=2 caption="Original residues and mutations in BCKDHA (1U5B chain A)">
 
<gallery widths=450px heights=250px perrow=2 caption="Original residues and mutations in BCKDHA (1U5B chain A)">
 
File:1u5b-orig.png|Original residues of BCKDHA are marked in <span style="color:red">'''red'''</span>.
 
File:1u5b-orig.png|Original residues of BCKDHA are marked in <span style="color:red">'''red'''</span>.
File:1u5b-mut.png|Mutant residues are marged in <span style="color:orange">'''orange'''</span>.
+
File:1u5b-mut.png|Mutant residues are marked in <span style="color:orange">'''orange'''</span>.
 
</gallery>
 
</gallery>
   
  +
<!--
 
=== Selected mutations ===
 
=== Selected mutations ===
  +
-->
 
Following are the selected mutations from HGMD and dbSNP. Changes in physicochemical properties, secondary structures, amino acid substitution scores, conservation and prediction results of different programs are also shown in the table:
 
 
{| class='wikitable' border='1' style='width:700px'
 
! mutation !! physicochemical changes !! secondary structure !! BLOSUM62 score !! PAM250 score !! PSSM conservation WT !! PSSM conservation mutant !! MSA conservation WT !! MSA conservation mutant !! SIFT !! Polyphen2 !! MutationTaster !! SNAP !! DB
 
|-
 
| P38H || small/fixed conformation/apolar to large/aromatic/polar || - || -2 || 0 || 31 || 0 || 0.29 || 0 || tolerated || benign || polymorphism || NA || dbSNP
 
|-
 
| M82L || sulfur-containing to aliphatic/branched || E || 2 || 4 || 12 || 59 || 0.36 || 0.01 || tolerated || benign || disease causing || NA || dbSNP
 
|-
 
| T151M || small/polar to apolar/sulfur-containing || - || -1 || -1 || 8 || 0 || 0.8 || 0 || tolerated || probably/possibly damaging || disease causing || NA || HGMD
 
|-
 
| A222T || apolar to polar || H || 0 || 1 || 64 || 4 || 0.38 || 0 || affect protein function || probably damaging || disease causing || NA || dbSNP
 
|-
 
| C264W || small/sulfur-containing to large/aromatic || E || -2 || -8 || 52 || 0 || 0.91 || 0 || affect protein function || probably/possibly damaging || disease causing || NA || HGMD
 
|-
 
| R265W || positively-charged to aromatic/larger || E || -3 || 2 || 12 || 0 || 0.36 || 0.38 || affect protein function || probably damaging || disease causing || NA || HGMD
 
|-
 
| R314Q || guanidin/large/positively-charged to amin || H || 1 || 1 || 23 || 0 || 0.35 || 0 || affect protein function || probably damaging || disease causing || NA || dbSNP
 
|-
 
| R346H || hydrophilic to hydrophobic/aromatic || - || 0 || 2 || 92 || 0 || 0.9 || 0 || affect protein function || probably damaging || disease causing || NA || HGMD
 
|-
 
| I361V || smaller || H || 3 || 4 || 64 || 10 || 0.89 || 0.01 || tolerated || benign/possibly damaging || disease causing || NA || dbSNP
 
|-
 
| Y413H || uncharged to positively-charged || - || 2 || 0 || 74 || 1 || 0.9 || 0 || affect protein function || probably damaging || disease causing || NA || HGMD
 
|}
 
   
 
=== SIFT ===
 
=== SIFT ===
Line 101: Line 77:
   
 
=== MutationTaster ===
 
=== MutationTaster ===
  +
Except the mutation P38H, all the other 9 selected mutations are predicted by MutationTaster to be disease causing.
  +
{| class='wikitable' border='1' style='width:700px'
  +
! mutation !! prediction !! mutation score !! prediction probability
  +
|- style="background-color: #AAAAFF"
  +
| P38H || polymorphism || 2.1 || 0.99
  +
|-
  +
| M82L || disease causing || 0.41 || 0.99
  +
|-
  +
| T151M || disease causing || 2.21 || 0.93
  +
|-
  +
| A222T || disease causing || 1.58 || 0.99
  +
|-
  +
| C264W || disease causing || 5.86 || 0.99
  +
|-
  +
| R265W || disease causing || 2.75 || 0.99
  +
|-
  +
| R314Q || disease causing || 1.17 || 0.99
  +
|-
  +
| I361V || disease causing || 0.79 || 0.99
  +
|-
  +
| R346H || disease causing || 0.79 || 0.99
  +
|-
  +
| Y413H || disease causing || 2.26 || 0.99
  +
|}
   
 
=== SNAP ===
 
=== SNAP ===
  +
  +
{| class='wikitable' border='1' style='width:700px'
  +
! Mutation !! Pathogenic !! RI !! Probability
  +
|-
  +
| P38H || Neutral || 4 || 72%
  +
|-
  +
| M82L || Neutral || 5 || 78%
  +
|- style="background-color: #ffaaaa;"
  +
| T151M || Non-neutral || 7 || 85%
  +
|- style="background-color: #ffaaaa;"
  +
| A222T || Non-neutral || 6 || 80%
  +
|- style="background-color: #ffaaaa;"
  +
| C264W || Non-neutral || 8 || 91%
  +
|- style="background-color: #ffaaaa;"
  +
| R265W || Non-neutral || 7 || 85%
  +
|-style="background-color: #ffaaaa;"
  +
| R314Q || Non-neutral || 6 || 80%
  +
|-style="background-color: #ffaaaa;"
  +
| R346H || Non-neutral || 4 || 71%
  +
|-
  +
| I361V || Neutral || 6 || 82%
  +
|-style="background-color: #ffaaaa;"
  +
| Y413H || Non-neutral || 8 || 91%
  +
|}
   
 
=== Summary ===
 
=== Summary ===
  +
  +
Following are the selected mutations from HGMD and dbSNP. Changes in physicochemical properties, secondary structures, amino acid substitution scores, conservation and prediction results of different programs are also shown in the table:
  +
  +
{| class='wikitable' border='1' style='width:700px'
  +
! mutation !! physicochemical changes !! secondary structure !! BLOSUM62 score !! PAM250 score !! PSSM conservation WT !! PSSM conservation mutant !! MSA conservation WT !! MSA conservation mutant !! SIFT !! Polyphen2 !! MutationTaster !! SNAP !! DB
  +
|-
  +
| P38H || small/fixed conformation/apolar to large/aromatic/polar || - || -2 || 0 || 0.31 || 0 || 0.29 || 0 || tolerated || benign || polymorphism || neutral || dbSNP
  +
|-
  +
| M82L || sulfur-containing to aliphatic/branched || E || 2 || 4 || 0.12 || 0.59 || 0.36 || 0.01 || tolerated || benign || disease causing || neutral || dbSNP
  +
|-
  +
| T151M || small/polar to apolar/sulfur-containing || - || -1 || -1 || 0.08 || 0 || 0.8 || 0 || tolerated || probably/possibly damaging || disease causing || non-neutral || HGMD
  +
|-
  +
| A222T || apolar to polar || H || 0 || 1 || 0.64 || 0.04 || 0.38 || 0 || affect protein function || probably damaging || disease causing || non-neutral || dbSNP
  +
|-
  +
| C264W || small/sulfur-containing to large/aromatic || E || -2 || -8 || 0.52 || 0 || 0.91 || 0 || affect protein function || probably/possibly damaging || disease causing || non-neutral || HGMD
  +
|-
  +
| R265W || positively-charged to aromatic/larger || E || -3 || 2 || 0.12 || 0 || 0.36 || 0.38 || affect protein function || probably damaging || disease causing || non-neutral || HGMD
  +
|-
  +
| R314Q || guanidin/large/positively-charged to amin || H || 1 || 1 || 0.23 || 0 || 0.35 || 0 || affect protein function || probably damaging || disease causing || non-neutral || dbSNP
  +
|-
  +
| R346H || hydrophilic to hydrophobic/aromatic || - || 0 || 2 || 0.92 || 0 || 0.90 || 0 || affect protein function || probably damaging || disease causing || non-neutral || HGMD
  +
|-
  +
| I361V || smaller || H || 3 || 4 || 0.64 || 0.10 || 0.89 || 0.01 || tolerated || benign/possibly damaging || disease causing || neutral || dbSNP
  +
|-
  +
| Y413H || uncharged to positively-charged || - || 2 || 0 || 0.74 || 0.01 || 0.90 || 0 || affect protein function || probably damaging || disease causing || non-neutral || HGMD
  +
|}
   
 
== Discussion ==
 
== Discussion ==
  +
  +
In six mutations (A222T, C264W, R265W, R314Q, R346H and Y413H) the prediction programs agree, that they could affect the protein function or be disease causing, respectively. Among these, there are the lowest BLOSUM62 and PAM250 scores but also higher scores. If the results of evolutionary conservation analysis by using a PSSM or a MSA are taken together, these mutations show a much higher conservation for the WT amino acid than for the mutant. They also show obvious changes in their physicochemical properties like change in charge, hydrophobicity or size. It is interesting that two of them (A222T and R314Q) are not included in HGMD, so it might be that they are disease causing but it is not known yet, or they are tolerated although there are these hints that they might change the protein function. The fact that the two mutations also lie in secondary structure elements argues for the first.
  +
  +
Only for the mutation P38H the programs agree that it has no effect. This is surprising because it has bad scores in the scoring matrices and also shows no conservation in the mutant amino acid. But because it lies near the end of the protein, the residue seems to have no crucial function, why it can be exchanged without effect. This view is also consistant with the mutation not beeing listed in HGMD.
  +
  +
For the remaining three muations (M82L, T151M and I361V), the programs come to different results, if it changes the protein function or not. For M82L and I361V it is unlikely, that they cause the disease, because they do not present high changes in the properties of the amino acids. Also the BLOSUM and PAM scores are comparatively high and the mutant amino acid is often observed in the conservation analysis - at least for the PSSM, which was derived from all homologs, but also for the MSA with mammalian homologs the mutant was still observed. But M82L and I361V also lie in secondary structure elements, so we cannot say definitely, that they are not disease causing. For T151M, the change of polarity, low BLOSUM and PAM scores and no conservation for the mutant amino acid argues for this mutation being disease causing - it actually is included in HGMD.
  +
  +
So in summary, most of the mutations that are known to cause MSUD and are consequently listed in HGMD could be identified with the predictions. For some mutations (e. g. A222T and R314Q) there are hints that they might affect the protein function, but at the moment it has not been published that they are disease causing. There is always an uncertainty, because if it has not been shown that a mutation is disease causing, does not mean that it is not disease causing. But the analyses have pointed out, that a mutation that introduces great changes in the physicochemical properties, has a high conservation for the WT and no conservation for the mutant amino acid and also lies in secondary structure elements, will likely change the function of the protein and therefore might lead to a disease.

Latest revision as of 15:49, 29 August 2013

Results

Lab journal

Visualization of mutations


SIFT

The following table shows the prediction of mutation effects by SIFT. 6 out of 10 mutations are predicted to affect protein function.


mutation prediction score median sequence conservation sequences represented at position
P38H tolerated 0.16 3.16 17
M82L tolerated 0.56 3.03 43
T151M tolerated 0.05 3.03 46
A222T affect protein function 0.04 3.03 46
C264W affect protein function 0.00 3.03 46
R265W affect protein function 0.00 3.03 46
R314Q affect protein function 0.01 3.03 46
R346H affect protein function 0.00 3.03 46
I361V tolerated 0.12 3.03 46
Y413H affect protein function 0.00 3.03 46


Polyphen2

The prediction resuls of Polyphen2 are shown in the following table. 7 out of 10 mutations are predicted to be probably damaging (at least for the trained set HumDiv). For the mutation T151M, this does not agree with the results of SIFT. But for this mutations the Polyphen scores are also the lowest of all mutations that are predicted to be damaging. So Polyphen sees an uncertainty for this mutation too, if it is damaging.


HumDiv HumVar
mutation prediction score sensitivity specificity prediction score sensitivity specificity
P38H benign 0.118 0.93 0.96 benign 0.042 0.93 0.62
M82L benign 0.002 0.99 0.30 benign 0.006 0.97 0.45
T151M probably damaging 0.963 0.78 0.95 possibly damaging 0.622 0.80 0.85
A222T probably damaging 1.000 0.00 1.00 probably damaging 0.943 0.65 0.91
C264W probably damaging 0.989 0.72 0.97 possibly damaging 0.887 0.71 0.89
R265W probably damaging 1.000 0.00 1.00 probably damaging 1.000 0.00 1.00
R314Q probably damaging 1.000 0.00 1.00 probably damaging 0.998 0.18 0.98
R346H probably damaging 1.000 0.00 1.00 probably damaging 0.993 0.47 0.96
I361V benign 0.386 0.90 0.89 possibly damaging 0.610 0.80 0.83
Y413H probably damaging 1.000 0.00 1.0 probably damaging 0.999 0.09 0.99

MutationTaster

Except the mutation P38H, all the other 9 selected mutations are predicted by MutationTaster to be disease causing.

mutation prediction mutation score prediction probability
P38H polymorphism 2.1 0.99
M82L disease causing 0.41 0.99
T151M disease causing 2.21 0.93
A222T disease causing 1.58 0.99
C264W disease causing 5.86 0.99
R265W disease causing 2.75 0.99
R314Q disease causing 1.17 0.99
I361V disease causing 0.79 0.99
R346H disease causing 0.79 0.99
Y413H disease causing 2.26 0.99

SNAP

Mutation Pathogenic RI Probability
P38H Neutral 4 72%
M82L Neutral 5 78%
T151M Non-neutral 7 85%
A222T Non-neutral 6 80%
C264W Non-neutral 8 91%
R265W Non-neutral 7 85%
R314Q Non-neutral 6 80%
R346H Non-neutral 4 71%
I361V Neutral 6 82%
Y413H Non-neutral 8 91%

Summary

Following are the selected mutations from HGMD and dbSNP. Changes in physicochemical properties, secondary structures, amino acid substitution scores, conservation and prediction results of different programs are also shown in the table:

mutation physicochemical changes secondary structure BLOSUM62 score PAM250 score PSSM conservation WT PSSM conservation mutant MSA conservation WT MSA conservation mutant SIFT Polyphen2 MutationTaster SNAP DB
P38H small/fixed conformation/apolar to large/aromatic/polar - -2 0 0.31 0 0.29 0 tolerated benign polymorphism neutral dbSNP
M82L sulfur-containing to aliphatic/branched E 2 4 0.12 0.59 0.36 0.01 tolerated benign disease causing neutral dbSNP
T151M small/polar to apolar/sulfur-containing - -1 -1 0.08 0 0.8 0 tolerated probably/possibly damaging disease causing non-neutral HGMD
A222T apolar to polar H 0 1 0.64 0.04 0.38 0 affect protein function probably damaging disease causing non-neutral dbSNP
C264W small/sulfur-containing to large/aromatic E -2 -8 0.52 0 0.91 0 affect protein function probably/possibly damaging disease causing non-neutral HGMD
R265W positively-charged to aromatic/larger E -3 2 0.12 0 0.36 0.38 affect protein function probably damaging disease causing non-neutral HGMD
R314Q guanidin/large/positively-charged to amin H 1 1 0.23 0 0.35 0 affect protein function probably damaging disease causing non-neutral dbSNP
R346H hydrophilic to hydrophobic/aromatic - 0 2 0.92 0 0.90 0 affect protein function probably damaging disease causing non-neutral HGMD
I361V smaller H 3 4 0.64 0.10 0.89 0.01 tolerated benign/possibly damaging disease causing neutral dbSNP
Y413H uncharged to positively-charged - 2 0 0.74 0.01 0.90 0 affect protein function probably damaging disease causing non-neutral HGMD

Discussion

In six mutations (A222T, C264W, R265W, R314Q, R346H and Y413H) the prediction programs agree, that they could affect the protein function or be disease causing, respectively. Among these, there are the lowest BLOSUM62 and PAM250 scores but also higher scores. If the results of evolutionary conservation analysis by using a PSSM or a MSA are taken together, these mutations show a much higher conservation for the WT amino acid than for the mutant. They also show obvious changes in their physicochemical properties like change in charge, hydrophobicity or size. It is interesting that two of them (A222T and R314Q) are not included in HGMD, so it might be that they are disease causing but it is not known yet, or they are tolerated although there are these hints that they might change the protein function. The fact that the two mutations also lie in secondary structure elements argues for the first.

Only for the mutation P38H the programs agree that it has no effect. This is surprising because it has bad scores in the scoring matrices and also shows no conservation in the mutant amino acid. But because it lies near the end of the protein, the residue seems to have no crucial function, why it can be exchanged without effect. This view is also consistant with the mutation not beeing listed in HGMD.

For the remaining three muations (M82L, T151M and I361V), the programs come to different results, if it changes the protein function or not. For M82L and I361V it is unlikely, that they cause the disease, because they do not present high changes in the properties of the amino acids. Also the BLOSUM and PAM scores are comparatively high and the mutant amino acid is often observed in the conservation analysis - at least for the PSSM, which was derived from all homologs, but also for the MSA with mammalian homologs the mutant was still observed. But M82L and I361V also lie in secondary structure elements, so we cannot say definitely, that they are not disease causing. For T151M, the change of polarity, low BLOSUM and PAM scores and no conservation for the mutant amino acid argues for this mutation being disease causing - it actually is included in HGMD.

So in summary, most of the mutations that are known to cause MSUD and are consequently listed in HGMD could be identified with the predictions. For some mutations (e. g. A222T and R314Q) there are hints that they might affect the protein function, but at the moment it has not been published that they are disease causing. There is always an uncertainty, because if it has not been shown that a mutation is disease causing, does not mean that it is not disease causing. But the analyses have pointed out, that a mutation that introduces great changes in the physicochemical properties, has a high conservation for the WT and no conservation for the mutant amino acid and also lies in secondary structure elements, will likely change the function of the protein and therefore might lead to a disease.