regSNP-intron is tool that predicts the disease-causing probability of intronic single nucleotide variants (iSNVs) based on both genomic and protein structural features.
The input of regSNP should be tab-delimited and contain four columns: chromosome, genomic coordniate, reference allele, alternative allele. The chromosome name should contain "chr" prefix (eg. chr1, chr2, etc.). The genomic coordinate is currently based on GRCh37/hg19. The reference/alternative allele should only contain single nucleotide.
The main result will show the probability that an iSNV can cause diseases. The iSNVs will then be classified into different categories basd on False Positive Rate (FPR). The full list of genomic and protein structural features will also be available for download.
| On splicing site | Off splicing site | |||||
|---|---|---|---|---|---|---|
| Category | Probability | FPR | TPR | Probability | FPR | TPR |
| Benign | [0.0, 0.36) | > 10% | > 89% | [0.0, 0.56) | > 10% | > 52% |
| Possibly Damaging | [0.36, 0.45) | <= 10% | <= 89% | [0.56, 0.61) | <= 10% | <= 52% |
| Damaging | [0.45, 1.0] | <= 5% | <= 84% | [0.61, 1.0] | <= 5% | <= 44% |
Please go to the submission page and make sure your job is completed. Also, only the intronic SNVs that are not overlapped with other exonic regions will be shown in the result. The Ensembl gene annotation we are using can be download here.