Difference between revisions of "Rs1800430"

From Bioinformatikpedia
(SNAP Prediction)
 
(29 intermediate revisions by 2 users not shown)
Line 20: Line 20:
 
== Sequence-based Mutation Analysis ==
 
== Sequence-based Mutation Analysis ==
   
=== Pysicochemical Properities ===
+
=== Pysicochemical Properties ===
   
 
First of all, we explored the amino acid properties and compared them for the original and the mutated amino acid. Therefore we created the possible effect that the mutation could have on the protein.
 
First of all, we explored the amino acid properties and compared them for the original and the mutated amino acid. Therefore we created the possible effect that the mutation could have on the protein.
Line 31: Line 31:
 
|polar, small, hydrophilic, negatively charged
 
|polar, small, hydrophilic, negatively charged
 
|polar, small, hydrophilic, negatively charged
 
|polar, small, hydrophilic, negatively charged
|Both amino acids have the same properities and therefore we suggest that an exchange of these two amino acids do not destroy the protein structure and function
+
|Both amino acids have the same properties and therefore we suggest that an exchange of these two amino acids do not destroy the protein structure and function
 
|-
 
|-
 
|}
 
|}
Line 39: Line 39:
 
----
 
----
   
=== Visualisation of the Mutation ===
+
=== Visualization of the Mutation ===
   
In the next step, we created the visualization of the muation with PyMol. Therefore we created a picture for the original amino acid, for the new mutated amino acid and finally for both together in one picture whereas the mutation is white colored. The following pictures display that the mutated amino acid Aspartic acid looks very different to Asparagine. They both agree in the first part of the rest but then go complety in opposite directions. Otherwise both rests fork at the end. All in all, the different directions prevail and will therefore probably cause big structural changes which will cause effects on the protein function.
+
In the next step, we created the visualization of the mutation with PyMol. Therefore we created a picture for the original amino acid (Figure 1), for the new mutated amino acid (Figure 2) and finally for both together in one picture whereas the mutation is white colored (Figure 3). The following pictures display that the mutated amino acid Aspartic acid looks very different to Asparagine. They both agree in the first part of the rest but then go completely in opposite directions. Otherwise both rests fork at the end. All in all, the different directions prevail and will therefore probably cause big structural changes which will cause effects on the protein function.
   
 
{| border="1" style="text-align:center; border-spacing:0;"
 
{| border="1" style="text-align:center; border-spacing:0;"
|picture original aa
+
|picture original amino acid
|picture mutated aa
+
|picture mutated amino acid
 
|combined picture
 
|combined picture
 
|-
 
|-
|[[Image:N399.png|thumb|150px|Amino acid Asparagine]]
+
|[[Image:N399.png|thumb|150px|Figure 1: Amino acid Asparagine]]
|[[Image:399D.png|thumb|150px|Amino acid Aspartic acid]]
+
|[[Image:399D.png|thumb|150px|Figure 2: Amino acid Aspartic acid]]
|[[Image:N399D.png|thumb|150px|Picture which visualize the mutation]]
+
|[[Image:N399D.png|thumb|150px|Figure 3: Picture which visualize the mutation]]
 
|-
 
|-
 
|}
 
|}
Line 58: Line 58:
 
----
 
----
   
=== Subsitution Matrices Values ===
+
=== Substitution Matrices Values ===
   
Afterwards, we looked at the values of the substitution matrices PAM1, PAM250 and BLOSSUM62. Therefore we looked detailed at the three values: the value for accoding amino acid substitution, the most frequent value for the substitution of the examined amino acid and the rarest substitution.
+
Afterwards, we looked at the values of the substitution matrices PAM1, PAM250 and BLOSSUM62. Therefore we looked detailed at the three values: the value for according amino acid substitution, the most frequent value for the substitution of the examined amino acid and the rarest substitution.
   
 
In this case, the substitution of Asparagine to Aspartic acid has very high values that agree with the most frequent value for all three substitution matrices. Therefore, according to all matrices a mutation at this position will probably not cause structural changes which can affect functional changes.
 
In this case, the substitution of Asparagine to Aspartic acid has very high values that agree with the most frequent value for all three substitution matrices. Therefore, according to all matrices a mutation at this position will probably not cause structural changes which can affect functional changes.
Line 69: Line 69:
 
|colspan="3" | BLOSOUM 62
 
|colspan="3" | BLOSOUM 62
 
|-
 
|-
|value aa
+
|value amino acid
 
|most frequent substitution
 
|most frequent substitution
 
|rarest substitution
 
|rarest substitution
|value aa
+
|value amino acid
 
|most frequent substitution
 
|most frequent substitution
 
|rarest substitution
 
|rarest substitution
|value aa
+
|value amino acid
 
|most frequent substitution
 
|most frequent substitution
 
|rarest substitution
 
|rarest substitution
Line 97: Line 97:
 
=== PSSM Analysis ===
 
=== PSSM Analysis ===
   
Besides, we looked additional at the position specific scoring matrix (PSSM) for ouer sequence. In contrast to PAM and BLOSOUM, the PSSM contains a specific substitution rate for each position in the sequence. Therefore, the PSSM is more position specific than PAM or BLOSOUM. We extracted the substitution value for the underlying mutation, the value for the most frequent substitution and the rarest substitution.
+
Besides, we looked additional at the position specific scoring matrix (PSSM) for our sequence. In contrast to PAM and BLOSOUM, the PSSM contains a specific substitution rate for each position in the sequence. Therefore, the PSSM is more position specific than PAM or BLOSOUM. We extracted the substitution value for the underlying mutation, the value for the most frequent substitution and the rarest substitution.
   
In this case the substitution rate for Asparagine to Aspartic acid at this position is in the middle between the value for the rarest substitution and the value for the most frequent substitution. Therefore, the value from PSSM is not realy significant and we are not able to determine effects on the protein.
+
In this case the substitution rate for Asparagine to Aspartic acid at this position is in the middle between the value for the rarest substitution and the value for the most frequent substitution. Therefore, the value from PSSM is not really significant and we are not able to determine effects on the protein.
   
 
{| border="1" style="text-align:center; border-spacing:0;"
 
{| border="1" style="text-align:center; border-spacing:0;"
 
|colspan="3" | PSSM
 
|colspan="3" | PSSM
 
|-
 
|-
|value aa
+
|value amino acid
 
|most frequent substitution
 
|most frequent substitution
 
|rarest substitution
 
|rarest substitution
Line 116: Line 116:
 
=== Conservation Analysis with Multiple Alignments ===
 
=== Conservation Analysis with Multiple Alignments ===
   
As a next step we created a multiple alignment which contains the HEXA sequence and 9 other mammalian homologous sequences from uniprot. Afterwards we looked at the position of the different mutations and looked at the conservation level on this position. The regarded mutation is presented by the first colored column. Here we can see, that almost all other mammalians have another amino acid or a gap on this position. Therefore, the mutation on this position is not highly conserved and a mutation there will cause probably no huge structural and functional changes for the protein.
+
As a next step we created a multiple alignment which contains the HEXA sequence and 9 other mammalian homologous sequences from [[http://www.uniprot.org UniProt]]. Afterwards we looked at the position of the different mutations and looked at the conservation level on this position. The regarded mutation is presented by the first colored column on Figure 4. Here we can see, that almost all other mammalians have another amino acid or a gap on this position. Therefore, the mutation on this position is not highly conserved and a mutation there will cause probably no huge structural and functional changes for the protein.
   
[[Image:mut_8.png|thumb|center|600px|Mutation in the multiple alignment]]
+
[[Image:mut_8.png|thumb|center|600px|Figure 4: Mutation in the multiple alignment]]
   
   
Line 126: Line 126:
 
=== Secondary Structure Mutation Analysis ===
 
=== Secondary Structure Mutation Analysis ===
   
As a next step we compared the different results of the secondary structure prediction tools JPred and PsiPred. Afterwards we can examine in which secondary structure element and where therein the mutation takes place. This can give an overview of how drastical the mutation can be.
+
As a next step we compared the different results of the secondary structure prediction tools JPred and PsiPred. Afterwards we can examine in which secondary structure element and where therein the mutation takes place. This can give an overview of how drastic the mutation can be.
 
In this case the tools disagree. JPred predict at the position of the mutation a coil and PsiPred predicted there a alpha-helix.
 
In this case the tools disagree. JPred predict at the position of the mutation a coil and PsiPred predicted there a alpha-helix.
For JPred a mutation at this position would not destroy or split a secondary structure element. It will probably only changes the coil between two secondary structure elements, but this can sometimes also cause a change of the the following secondary structure. We think that a drastical change of the protein structure and its function is unlikly for the JPred prediction.
+
For JPred a mutation at this position would not destroy or split a secondary structure element. It will probably only changes the coil between two secondary structure elements, but this can sometimes also cause a change of the the following secondary structure. We think that a drastic change of the protein structure and its function is unlikely for the JPred prediction.
 
For PsiPred a mutation at the beginning of the alpha helix could inhibit the formation if this alpha helix. Therefore, we think that a this mutation would probably cause a extreme change in the protein structure and function.
 
For PsiPred a mutation at the beginning of the alpha helix could inhibit the formation if this alpha helix. Therefore, we think that a this mutation would probably cause a extreme change in the protein structure and function.
   
Line 138: Line 138:
 
''' Comparison with the real Structure: '''
 
''' Comparison with the real Structure: '''
   
Afterwards we also visualize the position of the muation (red) in the real 3D-structure of PDB and compare it with the predicted secondary structure. The visualisation can therefore like above the predicted secondary structure display if the mutation is in a secondary structure element or in some other regions.
+
Afterwards we also visualize the position of the mutation (red) in the real 3D-structure of PDB and compare it with the predicted secondary structure (Figure 5 and Figure 6). The visualization can therefore like above the predicted secondary structure display if the mutation is in a secondary structure element or in some other regions.
   
Here in this case the mutationposition disagree with the position of the predicted secondary structure from JPred, but agrees with the position from PsiPred.
+
Here in this case the mutation position disagree with the position of the predicted secondary structure from JPred, but agrees with the position from PsiPred.
 
The visualization displays the mutation at the beginning of an alpha-helices. This means a mutation will probably cause structural changes combined with functional changes of the protein, because it can prevent the formation of this alpha-helix.
 
The visualization displays the mutation at the beginning of an alpha-helices. This means a mutation will probably cause structural changes combined with functional changes of the protein, because it can prevent the formation of this alpha-helix.
   
 
{|
 
{|
| [[Image:399_mut.png|thumb|250px|Mutation at position 399]]
+
| [[Image:399_mut.png|thumb|250px|Figure 5: Mutation at position 399]]
| [[Image:399_mut_detail.png|thumb|250px|Mutation at position 399 - detailed view]]
+
| [[Image:399_mut_detail.png|thumb|250px|Figure 6: Mutation at position 399 - detailed view]]
 
|}
 
|}
   
Line 154: Line 154:
 
=== SNAP Prediction ===
 
=== SNAP Prediction ===
   
Next, we looked at the result of the SNAP prediction. For this prediction we took the amino acid of the certain position and checked every possible amino acid mutation. Afterwards we extract the result for Aspartic acid which is the real mutation in this case. SNAP has a result that the exhange from Asparagine to Aspartic acid at this position is neutral with an average accuracy of 60%. This means that this certain mutation on this position cause likely no structural and functional changes of the protein.
+
Next, we looked at the result of the SNAP prediction. For this prediction we took the amino acid of the certain position and checked every possible amino acid mutation. Afterwards we extracted the result for Aspartic acid which is the real mutation in this case. SNAP has as a result that the exchange from Asparagine to Aspartic acid at this position is neutral with an average accuracy of 60%. This means that this certain mutation on this position cause likely no structural and functional changes of the protein.
   
 
{| border="1" style="text-align:center; border-spacing:0;"
 
{| border="1" style="text-align:center; border-spacing:0;"
Line 177: Line 177:
 
=== SIFT Prediction ===
 
=== SIFT Prediction ===
   
Next, we used SIFT Prediction which displays if a mutation is neutral or not. Therefore, it first shows a row which contains a score for the particular mutationposition to a certain amino acid. The amino acid which are not tolerated at this position are colored red. Besides, it also constructs a table which lists the amino acids that are predicted as tolerated and not-tolerated.
+
Next, we used SIFT Prediction which displays if a mutation is neutral or not. Therefore, it first shows a row which contains a score for the particular mutation position to a certain amino acid. The amino acid which are not tolerated at this position are colored red. Besides, it also constructs a table which lists the amino acids that are predicted as tolerated and not-tolerated.
   
In this case, there are five substitutions that are not tolerated: Tryptophan, Tyrosine, Phenylalanine, Cysteine and Methionine. The substitution to Aspartic acid is tolerated at this position. This means that this mutation at this position is probably neutral and will not cause any structural and function changes of the protein.
+
In this case, there are five substitutions that are not tolerated: Tryptophan, Tyrosine, Phenylalanine, Cysteine and Methionine. The substitution to Aspartic acid is tolerated at this position. This means that this mutation at this position is probably neutral and will not cause any structural and function changes of the protein (compare Figure 8).
   
 
SIFT Matrix:<br>
 
SIFT Matrix:<br>
Line 185: Line 185:
   
 
{|
 
{|
| [[Image:sift_legend.png|center]]
+
| [[Image:sift_legend.png|center|800px|thumb|Figure 7: Legend]]
 
|-
 
|-
| [[Image:399_sift.png.png|center]]
+
| [[Image:399_sift.png.png|center|800px|thumb|Figure 8: SIFT Table<br>
  +
Threshold for intolerance is 0.05.<BR>Amino acid color code: non-polar, <font color=green>uncharged polar</font>, <font color=red>basic</font>, <font color=blue>acidic</font>. <BR>Capital letters indicate amino acids appearing in the alignment, lower case letters result from prediction.]]
 
|}
 
|}
   
SIFT Table<br>
 
Threshold for intolerance is 0.05.<BR>Amino acid color code: nonpolar, <font color=green>uncharged polar</font>, <font color=red>basic</font>, <font color=blue>acidic</font>. <BR>Capital letters indicate amino acids appearing in the alignment, lower case letters result from prediction.
 
 
<br>
 
<br>
 
{| class="wikitable centered"
 
{| class="wikitable centered"
Line 205: Line 204:
 
=== PolyPhen2 Prediction ===
 
=== PolyPhen2 Prediction ===
   
Finally, we also regarded the PolyPhen2 prediction for this muation. This prediction visualizes have strongly demaging the mutation probably will be. Therefore it gives the result for two possible cases: HumDiv and HumVar. HumDiv is a prefered model for evaluation rare allels, dense mapping of regions identified by genome-wide assiociation studies and analysis of neutral selection. In contrast, HumVar is a prefered model for diagnostic of Mendelian diseases which require distinguishing mutations with drastic effects from all remaining human variations including abundant mildly deleterious allels. We decided to look at both possible models, which are agrees in the most cases.
+
Finally, we also regarded the PolyPhen2 prediction for this mutation. This prediction visualizes have strongly damaging the mutation probably will be. Therefore it gives the result for two possible cases: HumDiv and HumVar. HumDiv is the preferred model for evaluation rare alleles, dense mapping of regions identified by genome-wide association studies and analysis of neutral selection. In contrast, HumVar is the preferred model for diagnostic of Mendelian diseases which require distinguishing mutations with drastic effects from all remaining human variations including abundant mildly deleterious alleles. We decided to look at both possible models, which are agrees in the most cases.
   
In this case both models predict that the mutation is benign. This means that the mutation is neutral and will probably not damage the structure and the function of the protein.
+
In this case both models predict that the mutation is benign (Figure 9 and Figure 10). This means that the mutation is neutral and will probably not damage the structure and the function of the protein.
   
 
{|
 
{|
| [[Image:mut_8_humdiv.png|thumb|450px|HumDiv prediction]]
+
| [[Image:mut_8_humdiv.png|thumb|450px|Figure 9: HumDiv prediction]]
| [[Image:mut_8_humvar.png|thumb|450px|HumVar prediction]]
+
| [[Image:mut_8_humvar.png|thumb|450px|Figure 10: HumVar prediction]]
 
|}
 
|}
   
Line 218: Line 217:
   
 
== Structure-based Mutation Analysis ==
 
== Structure-based Mutation Analysis ==
  +
  +
=== Mapping onto Crystal Structure ===
  +
  +
[[Image:mut399_active.png|thumb|center|400px|Figure 11: Visualization of the mutation and important functional sites<br>Color declaration: <br>
  +
* <font color=red>red</font>: position of mutation<br>
  +
* <font color=green>green</font>: position of active side<br>
  +
* <font color=yellow>yellow</font>: position of glycolysation<br>
  +
* <font color=cyan>cyan</font>: position of Cysteine]]
  +
  +
First of all, we colored the important residues and also the mutated residue in the crystal structure, to see if the mutation is near of far away from the functional residues. As you can see on Figure 11, the mutation is located within a loop and far away from the functional residues. Therefore, we do not know in which way this mutation affects the global structure of the protein.
  +
  +
Back to [[http://i12r-studfilesrv.informatik.tu-muenchen.de/wiki/index.php/Structure-based_mutation_analysis_HEXA Structure-based mutation analysis]]
  +
----
   
 
=== SCWRL Prediction ===
 
=== SCWRL Prediction ===
   
  +
Because the mapping analysis does not give a good explanation why the mutation causes damages on the protein, we decided to analyse this mutation in more detail. Therefore, we looked for the structure of the original amino acid and the structure of the amino acid after the mutation event and compared them in size and orientation. For this purpose we used SCWRL.
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|picture original amino acid
  +
|picture mutated amino acid
  +
|combined picture
  +
|-
  +
|[[Image:mut399_org_aa.png|thumb|150px|Figure 12: Amino acid Asparagine]]
  +
|[[Image:mut399_aa.png|thumb|150px|Figure 13: Amino acid Aspartic acid]]
  +
|[[Image:mut399_both.png|thumb|150px|Figure 14: Picture which visualize the mutation]]
  +
|-
  +
|}
  +
  +
The pictures (Figure 12, Figure 13, Figure 14) show, that the structure of the different amino acids is very different. Although these two amino acids have very similar pysicochemical properties, they differ in their structure. The orientation of the residues is different and therefore it is possible, that these changes lead to missing or additional H-Bonds.
  +
  +
Back to [[http://i12r-studfilesrv.informatik.tu-muenchen.de/wiki/index.php/Structure-based_mutation_analysis_HEXA Structure-based mutation analysis]]
  +
----
  +
  +
=== FoldX Energy Comparison ===
  +
  +
Next we use the FoldX energy tool to compare the energy values of the two different structures.
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|Original total energy
  +
|Total energy for the mutated protein
  +
|Strongest energy changes within the mutated protein
  +
|-
  +
| -154.17
  +
| -155.11
  +
| -
  +
|-
  +
|}
  +
  +
The energy value of the mutated structure is lower than the energy value of the original structure. Therefore this means, that the structure is more rigid than the original structure. Although the difference is not dramatically it could lead to a too rigid structure which means that the binding site is too rigid to bind on its ligands.
  +
To have the possibility to compare these energy values with the values of the other analysis tools, we calculated a ratio between these energy values.
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|Ratio of the original structure
  +
|Ratio of the mutated structure
  +
|Difference
  +
|-
  +
|100
  +
|100.72
  +
| -0.72
  +
|-
  +
|}
  +
  +
Back to [[http://i12r-studfilesrv.informatik.tu-muenchen.de/wiki/index.php/Structure-based_mutation_analysis_HEXA Structure-based mutation analysis]]
  +
  +
----
  +
  +
=== Minimise Energy Comparison ===
  +
  +
Next we use the minimise energy tool to compare the energy values of the two different structures.
  +
  +
''' Comparing Energy: '''
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|Original total energy
  +
|Total energy for the mutated protein
  +
|-
  +
| -9610.467157
  +
| -9505.864181
  +
|-
  +
|}
  +
  +
In this case the energy value of the mutated structure is a little bit higher than the energy value of the original structure. This means that the mutated structure is not that rigid as the original one.
  +
To have the possibility to compare the energy values of the different tools we calculate the ratio.
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|Ratio of the original structure
  +
|Ratio of the mutated structure
  +
|Difference
  +
|-
  +
|100
  +
|98.91
  +
|1.09
  +
|-
  +
|}
  +
  +
With the ratio the difference between the energy of the original and mutated structure is about 1%.
  +
  +
''' Comparing Structure: '''
  +
  +
This tool also gives as output a [[http://www.pbd.org PDB]] file with the position of the original and the mutated amino acid.
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|picture original amino acid
  +
|picture mutated amino acid
  +
|combined picture
  +
|-
  +
|[[Image:mmut399_org.png|thumb|150px|Figure 15: Amino acid Asparagine]]
  +
|[[Image:mmut399_mut.png|thumb|150px|Figure 16: Amino acid Aspartic acid]]
  +
|[[Image:mmut399_both.png|thumb|150px|Figure 17: Picture which visualize the mutation]]
  +
|-
  +
|}
  +
  +
On the pictures (Figure 15, Figure 16, Figure 17) we can see, that the orientation of the residues is different. The result is similar to them of SCWRL (Figure 12, Figure 13, Figure 14).
  +
  +
''' Visualization of H-bonds and Clashes: '''
  +
  +
To get more insight in the effects of the mutated amino acid on the structure, we also analysed the H-bonds and clashes of the Serine residue.
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|H-bonds of the original structure
  +
|H-bonds near the mutation
  +
|Clashes of the mutation
  +
|-
  +
|[[Image:orig399.png|thumb|150px|Figure 18: H-bonds]]
  +
|[[Image:abc123.png|thumb|150px|Figure 19: H-bonds]]
  +
|[[Image:clash399.png|thumb|150px|Figure 20: Possible clashes]]
  +
|-
  +
|}
  +
  +
In both cases there are no H-bonds with other residues of the protein or with the backbone (Figure 18, Figure 19). Therefore, if this mutation lead to a loss of function of the protein it can not explained by missing or additional H-bonds. Furthermore, there are no clashes between the mutated amino acid and the rest of the protein (Figure 20).
  +
  +
Back to [[http://i12r-studfilesrv.informatik.tu-muenchen.de/wiki/index.php/Structure-based_mutation_analysis_HEXA Structure-based mutation analysis]]
  +
----
  +
  +
=== Gromacs Energy Comparison ===
  +
  +
''' Comparing Energy: '''
  +
  +
To analyse the energy values calculated by Gromacs, we used the AMBER99SB-ILDN force field.
  +
  +
Here are the values of the original structure:
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|Energy
  +
|Average
  +
|Err.Est.
  +
|RMSD
  +
|Tot-Drift
  +
|-
  +
|Bond
  +
|1091.57
  +
|270
  +
| -nan
  +
| -1622.75
  +
|-
  +
|Angle
  +
|3326.81
  +
|62
  +
| -nan
  +
|404.076
  +
|-
  +
|Potential
  +
| -61304.1
  +
|960
  +
| -nan
  +
| -6402.44
  +
|-
  +
|}
  +
  +
Here you can see the values which gromacs calculated for the structure with the mutated amino acid:
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|Energy
  +
|Average
  +
|Err.Est.
  +
|RMSD
  +
|Tot-Drift
  +
|-
  +
|Bond
  +
|938.527
  +
|210
  +
|2287.3
  +
| -1208.35
  +
|-
  +
|Angle
  +
|3201.09
  +
|59
  +
|240.737
  +
|389.204
  +
|-
  +
|Potential
  +
| -48693.9
  +
|870
  +
|3377.76
  +
| -5762.65
  +
|-
  +
|}
  +
  +
One difference between gromacs and the other tools we used is, that gromacs also calculated the energy for the bonds and the angles. To compare the energies between the different tools we only consider the potential energy in our analysis, because the potential energy is the energy of the complete protein. Therefore, we calculated the ratio between the energies only for the potential energy.
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|Ratio of the original structure
  +
|Ratio of the mutated structure
  +
|Difference
  +
|-
  +
|100
  +
|79.43
  +
|20.57
  +
|-
  +
|}
  +
  +
The result of Gromacs is consistent with the result of minimise. In this case, the original structure has a lower energy value than the mutated structure, which means that the mutated structure is more stable than the mutated structure.
  +
  +
''' Comparing Structure: '''
  +
  +
Gromacs also offers pictures of the mutated amino acids which can be seen in the following section.
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|picture original amino acid
  +
|picture mutated amino acid
  +
|combined picture
  +
|-
  +
|[[Image:gro_mut399_org.png|thumb|150px|Figure 21: Amino acid Asparagine]]
  +
|[[Image:gro_mut399_mut.png|thumb|150px|Figure 22: Amino acid Aspartic acid]]
  +
|[[Image:gro_mut399_both.png|thumb|150px|Figure 23: Picture which visualize the mutation]]
  +
|-
  +
|}
  +
  +
The three pictures (Figure 21, Figure 22, Figure 23) are similar to the result of SCWRL and minimise.
  +
  +
''' Visualization of H-bonds and Clashes: '''
  +
  +
{| border="1" style="text-align:center; border-spacing:0;"
  +
|H-bonds of the original structure
  +
|H-bonds near the mutation
  +
|Clashes of the mutation
  +
|-
  +
|[[Image:orig399.png|thumb|150px|Figure 24: H-bonds]]
  +
|[[Image:gro_hbond399.png|thumb|150px|Figure 25: H-bonds]]
  +
|[[Image:gro_clash399.png|thumb|150px|Figure 26: Possible clashes]]
  +
|-
  +
|}
   
  +
Neither the original (Figure 24) nor the mutated structure (Figure 25) have any H-bonds with the rest of the protein. The mutation also do not cause any clashes with the rest of the protein (Figure 26).
   
Back to [[http://i12r-studfilesrv.informatik.tu-muenchen.de/wiki/index.php/Structure-based_mutation_analysis_HEXA Sequence-based mutation analysis]]
+
Back to [[http://i12r-studfilesrv.informatik.tu-muenchen.de/wiki/index.php/Structure-based_mutation_analysis_HEXA Structure-based mutation analysis]]
 
----
 
----

Latest revision as of 22:41, 31 August 2011

General Information

SNP-id rs1800430
Codon 399
Mutation Codon Asn -> Asp
Mutation Triplet AAC -> GAC

Back to [Sequence-based mutation analysis]

Sequence-based Mutation Analysis

Pysicochemical Properties

First of all, we explored the amino acid properties and compared them for the original and the mutated amino acid. Therefore we created the possible effect that the mutation could have on the protein.

Asn Asp consequences
polar, small, hydrophilic, negatively charged polar, small, hydrophilic, negatively charged Both amino acids have the same properties and therefore we suggest that an exchange of these two amino acids do not destroy the protein structure and function


Back to [Sequence-based mutation analysis]


Visualization of the Mutation

In the next step, we created the visualization of the mutation with PyMol. Therefore we created a picture for the original amino acid (Figure 1), for the new mutated amino acid (Figure 2) and finally for both together in one picture whereas the mutation is white colored (Figure 3). The following pictures display that the mutated amino acid Aspartic acid looks very different to Asparagine. They both agree in the first part of the rest but then go completely in opposite directions. Otherwise both rests fork at the end. All in all, the different directions prevail and will therefore probably cause big structural changes which will cause effects on the protein function.

picture original amino acid picture mutated amino acid combined picture
Figure 1: Amino acid Asparagine
Figure 2: Amino acid Aspartic acid
Figure 3: Picture which visualize the mutation


Back to [Sequence-based mutation analysis]


Substitution Matrices Values

Afterwards, we looked at the values of the substitution matrices PAM1, PAM250 and BLOSSUM62. Therefore we looked detailed at the three values: the value for according amino acid substitution, the most frequent value for the substitution of the examined amino acid and the rarest substitution.

In this case, the substitution of Asparagine to Aspartic acid has very high values that agree with the most frequent value for all three substitution matrices. Therefore, according to all matrices a mutation at this position will probably not cause structural changes which can affect functional changes.

PAM 1 Pam 250 BLOSOUM 62
value amino acid most frequent substitution rarest substitution value amino acid most frequent substitution rarest substitution value amino acid most frequent substitution rarest substitution
36 36 (Asp) 0 (Cys, Met) 7 7 (Asp) 2 (Cys, Leu, Phe, Trp) 1 1 (Asp, His, Ser) -4 (Trp)


Back to [Sequence-based mutation analysis]


PSSM Analysis

Besides, we looked additional at the position specific scoring matrix (PSSM) for our sequence. In contrast to PAM and BLOSOUM, the PSSM contains a specific substitution rate for each position in the sequence. Therefore, the PSSM is more position specific than PAM or BLOSOUM. We extracted the substitution value for the underlying mutation, the value for the most frequent substitution and the rarest substitution.

In this case the substitution rate for Asparagine to Aspartic acid at this position is in the middle between the value for the rarest substitution and the value for the most frequent substitution. Therefore, the value from PSSM is not really significant and we are not able to determine effects on the protein.

PSSM
value amino acid most frequent substitution rarest substitution
0 2 -2

Conservation Analysis with Multiple Alignments

As a next step we created a multiple alignment which contains the HEXA sequence and 9 other mammalian homologous sequences from [UniProt]. Afterwards we looked at the position of the different mutations and looked at the conservation level on this position. The regarded mutation is presented by the first colored column on Figure 4. Here we can see, that almost all other mammalians have another amino acid or a gap on this position. Therefore, the mutation on this position is not highly conserved and a mutation there will cause probably no huge structural and functional changes for the protein.

Figure 4: Mutation in the multiple alignment


Back to [Sequence-based mutation analysis]


Secondary Structure Mutation Analysis

As a next step we compared the different results of the secondary structure prediction tools JPred and PsiPred. Afterwards we can examine in which secondary structure element and where therein the mutation takes place. This can give an overview of how drastic the mutation can be. In this case the tools disagree. JPred predict at the position of the mutation a coil and PsiPred predicted there a alpha-helix. For JPred a mutation at this position would not destroy or split a secondary structure element. It will probably only changes the coil between two secondary structure elements, but this can sometimes also cause a change of the the following secondary structure. We think that a drastic change of the protein structure and its function is unlikely for the JPred prediction. For PsiPred a mutation at the beginning of the alpha helix could inhibit the formation if this alpha helix. Therefore, we think that a this mutation would probably cause a extreme change in the protein structure and function.

JPred:
...CCCCCCCEEEEEECCCCCCCCHHHHHHHHHCCCCEEECCCCCCCCCCCCCCCCCC...
PsiPred:
...CCCCCCCCEEEEECCCCCCCHHHHHHHHHHCCCEEEECCCCCCCCCCCCCCCCCC...

Comparison with the real Structure:

Afterwards we also visualize the position of the mutation (red) in the real 3D-structure of PDB and compare it with the predicted secondary structure (Figure 5 and Figure 6). The visualization can therefore like above the predicted secondary structure display if the mutation is in a secondary structure element or in some other regions.

Here in this case the mutation position disagree with the position of the predicted secondary structure from JPred, but agrees with the position from PsiPred. The visualization displays the mutation at the beginning of an alpha-helices. This means a mutation will probably cause structural changes combined with functional changes of the protein, because it can prevent the formation of this alpha-helix.

Figure 5: Mutation at position 399
Figure 6: Mutation at position 399 - detailed view


Back to [Sequence-based mutation analysis]


SNAP Prediction

Next, we looked at the result of the SNAP prediction. For this prediction we took the amino acid of the certain position and checked every possible amino acid mutation. Afterwards we extracted the result for Aspartic acid which is the real mutation in this case. SNAP has as a result that the exchange from Asparagine to Aspartic acid at this position is neutral with an average accuracy of 60%. This means that this certain mutation on this position cause likely no structural and functional changes of the protein.

Substitution Prediction Reliability Index Expected Accuracy
D Neutral 1 60%

A detailed list of all possible substitutions can be found [here]


Back to [Sequence-based mutation analysis]


SIFT Prediction

Next, we used SIFT Prediction which displays if a mutation is neutral or not. Therefore, it first shows a row which contains a score for the particular mutation position to a certain amino acid. The amino acid which are not tolerated at this position are colored red. Besides, it also constructs a table which lists the amino acids that are predicted as tolerated and not-tolerated.

In this case, there are five substitutions that are not tolerated: Tryptophan, Tyrosine, Phenylalanine, Cysteine and Methionine. The substitution to Aspartic acid is tolerated at this position. This means that this mutation at this position is probably neutral and will not cause any structural and function changes of the protein (compare Figure 8).

SIFT Matrix:
Each entry contains the score at a particular position (row) for an amino acid substitution (column). Substitutions predicted to be intolerant are highlighted in red.

Figure 7: Legend
Figure 8: SIFT Table
Threshold for intolerance is 0.05.
Amino acid color code: non-polar, uncharged polar, basic, acidic.
Capital letters indicate amino acids appearing in the alignment, lower case letters result from prediction.




Predict Not ToleratedPositionSeq RepPredict Tolerated
wyfcm399N0.72iHvlgPtRDEQASKN




Back to [Sequence-based mutation analysis]


PolyPhen2 Prediction

Finally, we also regarded the PolyPhen2 prediction for this mutation. This prediction visualizes have strongly damaging the mutation probably will be. Therefore it gives the result for two possible cases: HumDiv and HumVar. HumDiv is the preferred model for evaluation rare alleles, dense mapping of regions identified by genome-wide association studies and analysis of neutral selection. In contrast, HumVar is the preferred model for diagnostic of Mendelian diseases which require distinguishing mutations with drastic effects from all remaining human variations including abundant mildly deleterious alleles. We decided to look at both possible models, which are agrees in the most cases.

In this case both models predict that the mutation is benign (Figure 9 and Figure 10). This means that the mutation is neutral and will probably not damage the structure and the function of the protein.

Figure 9: HumDiv prediction
Figure 10: HumVar prediction


Back to [Sequence-based mutation analysis]

Structure-based Mutation Analysis

Mapping onto Crystal Structure

Figure 11: Visualization of the mutation and important functional sites
Color declaration:
* red: position of mutation
* green: position of active side
* yellow: position of glycolysation
* cyan: position of Cysteine

First of all, we colored the important residues and also the mutated residue in the crystal structure, to see if the mutation is near of far away from the functional residues. As you can see on Figure 11, the mutation is located within a loop and far away from the functional residues. Therefore, we do not know in which way this mutation affects the global structure of the protein.

Back to [Structure-based mutation analysis]


SCWRL Prediction

Because the mapping analysis does not give a good explanation why the mutation causes damages on the protein, we decided to analyse this mutation in more detail. Therefore, we looked for the structure of the original amino acid and the structure of the amino acid after the mutation event and compared them in size and orientation. For this purpose we used SCWRL.

picture original amino acid picture mutated amino acid combined picture
Figure 12: Amino acid Asparagine
Figure 13: Amino acid Aspartic acid
Figure 14: Picture which visualize the mutation

The pictures (Figure 12, Figure 13, Figure 14) show, that the structure of the different amino acids is very different. Although these two amino acids have very similar pysicochemical properties, they differ in their structure. The orientation of the residues is different and therefore it is possible, that these changes lead to missing or additional H-Bonds.

Back to [Structure-based mutation analysis]


FoldX Energy Comparison

Next we use the FoldX energy tool to compare the energy values of the two different structures.

Original total energy Total energy for the mutated protein Strongest energy changes within the mutated protein
-154.17 -155.11 -

The energy value of the mutated structure is lower than the energy value of the original structure. Therefore this means, that the structure is more rigid than the original structure. Although the difference is not dramatically it could lead to a too rigid structure which means that the binding site is too rigid to bind on its ligands. To have the possibility to compare these energy values with the values of the other analysis tools, we calculated a ratio between these energy values.

Ratio of the original structure Ratio of the mutated structure Difference
100 100.72 -0.72

Back to [Structure-based mutation analysis]


Minimise Energy Comparison

Next we use the minimise energy tool to compare the energy values of the two different structures.

Comparing Energy:

Original total energy Total energy for the mutated protein
-9610.467157 -9505.864181

In this case the energy value of the mutated structure is a little bit higher than the energy value of the original structure. This means that the mutated structure is not that rigid as the original one. To have the possibility to compare the energy values of the different tools we calculate the ratio.

Ratio of the original structure Ratio of the mutated structure Difference
100 98.91 1.09

With the ratio the difference between the energy of the original and mutated structure is about 1%.

Comparing Structure:

This tool also gives as output a [PDB] file with the position of the original and the mutated amino acid.

picture original amino acid picture mutated amino acid combined picture
Figure 15: Amino acid Asparagine
Figure 16: Amino acid Aspartic acid
Figure 17: Picture which visualize the mutation

On the pictures (Figure 15, Figure 16, Figure 17) we can see, that the orientation of the residues is different. The result is similar to them of SCWRL (Figure 12, Figure 13, Figure 14).

Visualization of H-bonds and Clashes:

To get more insight in the effects of the mutated amino acid on the structure, we also analysed the H-bonds and clashes of the Serine residue.

H-bonds of the original structure H-bonds near the mutation Clashes of the mutation
Figure 18: H-bonds
Figure 19: H-bonds
Figure 20: Possible clashes

In both cases there are no H-bonds with other residues of the protein or with the backbone (Figure 18, Figure 19). Therefore, if this mutation lead to a loss of function of the protein it can not explained by missing or additional H-bonds. Furthermore, there are no clashes between the mutated amino acid and the rest of the protein (Figure 20).

Back to [Structure-based mutation analysis]


Gromacs Energy Comparison

Comparing Energy:

To analyse the energy values calculated by Gromacs, we used the AMBER99SB-ILDN force field.

Here are the values of the original structure:

Energy Average Err.Est. RMSD Tot-Drift
Bond 1091.57 270 -nan -1622.75
Angle 3326.81 62 -nan 404.076
Potential -61304.1 960 -nan -6402.44

Here you can see the values which gromacs calculated for the structure with the mutated amino acid:

Energy Average Err.Est. RMSD Tot-Drift
Bond 938.527 210 2287.3 -1208.35
Angle 3201.09 59 240.737 389.204
Potential -48693.9 870 3377.76 -5762.65

One difference between gromacs and the other tools we used is, that gromacs also calculated the energy for the bonds and the angles. To compare the energies between the different tools we only consider the potential energy in our analysis, because the potential energy is the energy of the complete protein. Therefore, we calculated the ratio between the energies only for the potential energy.

Ratio of the original structure Ratio of the mutated structure Difference
100 79.43 20.57

The result of Gromacs is consistent with the result of minimise. In this case, the original structure has a lower energy value than the mutated structure, which means that the mutated structure is more stable than the mutated structure.

Comparing Structure:

Gromacs also offers pictures of the mutated amino acids which can be seen in the following section.

picture original amino acid picture mutated amino acid combined picture
Figure 21: Amino acid Asparagine
Figure 22: Amino acid Aspartic acid
Figure 23: Picture which visualize the mutation

The three pictures (Figure 21, Figure 22, Figure 23) are similar to the result of SCWRL and minimise.

Visualization of H-bonds and Clashes:

H-bonds of the original structure H-bonds near the mutation Clashes of the mutation
Figure 24: H-bonds
Figure 25: H-bonds
Figure 26: Possible clashes

Neither the original (Figure 24) nor the mutated structure (Figure 25) have any H-bonds with the rest of the protein. The mutation also do not cause any clashes with the rest of the protein (Figure 26).

Back to [Structure-based mutation analysis]