pBRIT candiate gene prioritization

contact

Prioritize Gene Lists

pBRIT prioritizes candidate genes based on a set of training genes provided by the user. pBRIT incorporates two information-theoretic approaches for data fusion:

  • TFIDF
  • TFIDF_SVD

Under each of the data fusion methods, there are two alternative regression designs:

  • Test.Pheno.Include : Phenotypic information of the candidate genes is included in the regression model.
  • Test.Pheno.Discard : Phenotypic information of the candidate genes is not included in the regression model.

In our cross-validation studies, TFIDF_SVD combined with Test.Pheno.Include gave comparatively better results.



Set Run Parameters

: Provide a name to identify your prioritization job.

: Optionally provide your email to recieve a notification when the prioritization is finished.

: pBRIT datafusion method

: Include known test-gene phenotype annotations into the regression model. Discarding the phenotype information forces pBRIT to purley predict the phenotype association on the functional annotations.

: Database Release. The publication was based on release January 2015.

Training GenesTest Genes
A set of genes known to be related to the phenotype of the patient. These genes will be used to train the model. Please provide at least "3" Training genes. A set of genes affected in the patient. These genes will be ranked on their similarity to the training set.
Required format: GeneIDs according to HUGO, one per line. Required format: GeneIDs according to HUGO, one per line.