**********************************************************************************************
**                                                                                          **
**                                         SNPs&GO                                          **
**                  Predicting disease associated variation using GO terms                  **
**                                                                                          **
**********************************************************************************************

      Sequence File: fileseq.seq

      Mutation     Prediction   RI  Probability   Method
          R63H        Neutral    0        0.488   PANTHER:  F[R]=41%  F[H]=2%  
          R63H        Neutral    3        0.368   SNPs&GO
         E322K        Neutral    1        0.470   PANTHER:  F[E]=62%  F[K]=3%  
         E322K        Disease    3        0.663   SNPs&GO
         R337Q        Disease    1        0.529   PANTHER:  F[R]=65%  F[Q]=2%  
         R337Q        Disease    0        0.522   SNPs&GO
         M350I        Neutral    6        0.178   PANTHER:  F[M]=32%  F[I]=8%  
         M350I        Neutral    4        0.289   SNPs&GO
         G439V        Disease   10        0.987   PANTHER:  F[G]=96%  F[V]=0%  
         G439V        Disease    7        0.836   SNPs&GO
         D470N   Unclassified   NA           NA   PANTHER:  F[D]=NA   F[N]=NA
         D470N        Neutral    6        0.186   SNPs&GO

      Mutation: WT+POS+NEW
      WT:  Residue in wild-type protein
      POS: Residue position
      NEW: New residue after mutation
      Prediction:
             Neutral: Neutral variation
             Disease: Disease associated variation
      RI: Reliability Index
      Probability: Disease probability (if >0.5 mutation is predicted Disease)
      Method: SVM type and data
             PANTHER: Output of the PANTHER algorithm
             PhD-SNP:  SVM input is the sequence and profile at the mutated position
             SNPs&GO:  SVM input is all the input in PhD-SNP, PANTHER  and GO term features
      F[X]: Frequency of residue X in the sequence profile
      Nali: Number of aligned sequences in the mutated site

**********************************************************************************************
**                                                                                          **
**     Calabrese R, Capriotti E, Fariselli P, Martelli PL, Casadio R. (2009). Functional    **
**     annotations improve the predictive score of human disease-related mutations in       **
**     proteins. Human Mutation. 30:1237-1244.                                              **
**                                                                                          **
**     Capriotti E, Altman RB. (2011). Improving the prediction of disease-related vari-    **
**     ants using protein three-dimensional structure. BMC Bioinformatics. 12 (Sup.4) S3.   **
**                                                                                          **
**********************************************************************************************