File S1 - Prediction of Disease Causing Non-Synonymous SNPs by the Artificial Neural Network Predictor NetDiseaseSNP

This file contains three supporting tables. Table S1aa Performance of NetDiseaseSNP. (1) All SNPs where NetDiseaseSNP and SIFT agree on the prediction; (2) All SNPs where NetDiseaseSNP and SIFT disagree on the prediction; (3) All SNPs where SIFT is not able to generate a prediction; (4) All SNPs. Table S1ab Performance of SIFT. (1) All SNPs where NetDiseaseSNP and SIFT agree on the prediction; (2) All SNPs where NetDiseaseSNP and SIFT disagree on the prediction; (3) All SNPs where SIFT can generate a prediction. Table S1ba Performance of NetDiseaseSNP. (1) All SIFT data encoded SNPs where NetDiseaseSNP and SIFT agree on the prediction; (2) All SIFT data encoded SNPs where NetDiseaseSNP and SIFT disagree on the prediction; (3) All SIFT data encoded SNPs. Table S1bb Performance of SIFT. (1) All SIFT data encoded SNPs where NetDiseaseSNP and SIFT agree on the prediction; (2) All SIFT data encoded SNPs where NetDiseaseSNP and SIFT disagree on the prediction; (3) All SIFT data encoded SNPs where SIFT can generate a prediction. Table S1ca Performance of NetDiseaseSNP. (1) All Blosum62 data encoded SNPs where NetDiseaseSNP and SIFT agree on the prediction; (2) All Blosum62 data encoded SNPs where NetDiseaseSNP and SIFT disagree on the prediction; (3) All Blosum62 data encoded SNPs where SIFT is not able to generate a prediction; (4) All Blosum62 data encoded SNPs. Table S1cb Performance of SIFT. (1) All Blosum62 data encoded SNPs where NetDiseaseSNP and SIFT agree on the prediction; (2) All Blosum62 data encoded SNPs where NetDiseaseSNP and SIFT disagree on the prediction; (3) All Blosum62 data encoded SNPs where SIFT can generate a prediction. Table S1da Performance of NetDiseaseSNP. (1) All Blosum62 data encoded SNPs where SIFT predictions exist and NetDiseaseSNP and SIFT agree on the prediction; (2) All Blosum62 data encoded SNPs where SIFT predictions exist and NetDiseaseSNP and SIFT disagree on the prediction; (3) All Blosum62 data encoded SNPs where SIFT predictions exist. Table S1db Performance of SIFT. (1) All Blosum62 data encoded SNPs where SIFT predictions exist and NetDiseaseSNP and SIFT agree on the prediction; (2) All Blosum62 data encoded SNPs where SIFT predictions exist and NetDiseaseSNP and SIFT disagree on the prediction; (3) All Blosum62 data encoded SNPs where SIFT predictions exist and SIFT can generate a prediction. Table S1e Number of neutral and disease SNPs for each of different types of encoding for SNPs. The columns in the table are: Column 1: Input data to NetDiseaseSNP; Column 2: Protein is longer than 2000 amino acids; Column 3: SIFT data exists for the all SNPs; Column 4: Number of neutral SNPs; Column 5: Number of disease SNPs. The rows in the table are: (1) All SNPs both SIFT and Blosum62 data encoded SNPs; (2) SIFT data encoded SNPs; (3) Blosum62 data encoded SNPs; (4) Blosum62 data encoded SNPs where SIFT output data exists i.e. protein is longer than 2000 amino acids; (5) Blosum62 data encoded SNPs where SIFT output data does not exist. Table S2a Predictions by NetDiseaseSNP and SIFT. The rows in the table are: (1) Both NetDiseaseSNP and SIFT predict disease; (2) Both NetDiseaseSNP and SIFT predict neutral; (3) NetDiseaseSNP predicts disease and SIFT predicts neutral; (4) NetDiseaseSNP predicts neutral and SIFT predicts disease; (5) NetDiseaseSNP predicts disease and SIFT no prediction; (6) NetDiseaseSNP predicts neutral and SIFT no prediction; (7) Total number of mutations. Table S2b Predictions by NetDiseaseSNP on mutations where NetDiseaseSNP predicts the mutation to be a disease mutation and mutations are encoded with Blosum62 matrix data. The columns in the table are: Column 1: Description of the data in the row; Column 2: Protein is longer than 2000 amino acids; Column 3: All mutations at all positions in the protein are encoded with Blosum62 matrix data. The rows in the table are: (1) Protein longer than 2000 amino acids; (2) SIFT is not able to generate output for any mutation in this protein and the protein is shorter than 2000 amino acids; (3) SIFT is able to generate output for some mutations in this protein and the protein is shorter than 2000 amino acids; (4) Total number of mutations. Table S2c Predictions by NetDiseaseSNP on mutations where NetDiseaseSNP predicts the mutation to be a neutral mutation and mutations are encoded with Blosum62 matrix data. The columns in the table are: Column 1: Description of the data in the row; Column 2: Protein is longer than 2000 amino acids; Column 3: All mutations at all positions in the protein are encoded with Blosum62 matrix data. The rows in the table are: (1) Protein longer than 2000 amino acids; (2) SIFT is not able to generate output for any mutation in this protein and the protein is shorter than 2000 amino acids; (3) SIFT is able to generate output for some mutations in this protein and the protein is shorter than 2000 amino acids; (4) Total number of mutations. Table S3a Performance of NetDiseaseSNP. (1) All mutations where NetDiseaseSNP and SIFT agree on the prediction; (2) All mutations where NetDiseaseSNP and SIFT disagree on the prediction; (3) All mutations where SIFT is not able to generate a prediction; (4) All mutations. Table S3b Performance of SIFT. (1) All mutations where NetDiseaseSNP and SIFT agree on the prediction; (2) All mutations where NetDiseaseSNP and SIFT disagree on the prediction; (3) All mutations where SIFT can generate a prediction.
(DOC)