MutaSense

Fechter K, Porollo A. MutaCYP: Classification of missense mutations in human cytochromes P450. BMC Med Genomics. 2014 Jul 30;7:47.

Training set TS270 (Table 1S)
Blind set BS292 (Table 2S)
Description of features and their F-scores (Table 3S)
Feature selection (Table 4S)
Neural network model selection (Table 5S)
Consensus based models (Table 6S)

Currently, MutaCYP is the only option for functional annotation of missense mutations provided by MutaSense. As it follows from the name of the method, MutaCYP is limited to predictions in human cytochromes P450 (CYPs).

Input

The server expects a list of mutations, one per line, in either of the two following formats:

1. rsNumber, where Number is a dbSNP identifier (Reference SNP ID)

Example:

rs6161
rs2229381
rs9282671
rs28934586
rs59443548
    

2. UniProtID AArefPosAAmut, where
UniProtID − UniProt ID of a protein, which can be entered by a double mouse click from the list provided on the submission form;
AAref − amino acid in the reference sequence
Pos − position of the mutation
AAmut − amino acid resulting from the mutation.
For multiple mutations at the same position, they can be specified in one line (see example below for position R474).

Example:

P05108 L141W
P05108 A189V
P05108 L222P
P05108 E314K
P05108 R353W
P05108 A359V
P05108 V415E
Q02318 G145E
Q02318 A169V
Q02318 T175M
Q02318 R395C
Q02318 R395S
Q02318 R405Q
Q02318 R474QW
Q02318 R479C
    

Output

MutaCYP makes prediction as to whether a given mutation is deleterious or benign using a neural network-based prediction model considering 5 sequence-based features. The method provides a score from 0 to 1, with a deleterious mutation being assigned to the score ≥ 0.5. All mutations for a given CYP can be mapped to either a sequence-based view of a protein (using POLYVIEW-2D) or to a 3D interactive view (using the Jmol applet). Additionally, characteristics of each mutation can be compared to the distributions of the corresponding characteristics of 270 known benign and deleterious mutations in CYPs.

Example of prediction results for the P05108 L141W mutation.

Using boxplots

Using frequency histograms

Using POLYVIEW-2D sequence profiles

Amino acid sequence aligned with secondary structure elements, relative solvent accessibility, and CYP structural annotation.

CYP structural annotation codes:
X − a residue is not available in 3D structure
- − no special annotation available
C − a residue is located in the active site cavity
H − a residue is in contact with heme
S − a residue is in contact with a known substrate
D − a residue is in contact with both heme and a known substrate

Other annotations can be found in the POLYVIEW-2D documentation

Amino acid sequence with position specific conservation scores mapped.

3D view of CYP with the mutated residue rendered using sticks and labeled (numbering corresponds to the PDB coordinate file).