deTiN

What is deTiN
DeTiN is designed to measure tumor in normal (TiN) contamination and improve sensitivity by rescuing rejected sites when using a contaminated matched control.

For complete details please see our publication: (pending)

How does it work

In brief deTiN works in two steps:
1) Estimating TiN: DeTiN models the matched normal sample as a mixture of normal cells and an unknown fraction of contaminating cells from the matched tumor sample. The TiN estimate should be interpreted as a relative fraction of tumor cells in the tumor and normal samples. The method uses two independent types of tumor-specific events: (i) candidate somatic single nucleotide variations (SSNVs); and (ii) genomic regions of allelic imbalance in the tumor (including deletions, amplifications, and copy neutral loss-of-heterozygosity), extracted from estimated allele-specific somatic copy number alterations (aSCNAs) found in the tumor . We calculate the posterior distribution for TiN values based on each of the two types of somatic events separately and then combine them to identify the maximum a posteriori (MAP) value (and a confidence interval). We flag cases in which the TiN estimates based on each data type separately are inconsistent, potentially identifying multi-clonal tumor in normal.

2) Recovering somatic variants: Next, we use the estimated TiN to recover likely somatic mutations (both SSNVs and indels). Briefly, for each candidate variant, deTiN probabilistically compares two scenarios. In the first scenario, the alternate allele count in the normal sample represents an underlying germline variant; and in the second scenario, it represents a somatic variant coming from tumor cells mixed at the estimated TiN value.  The method takes into account the prior probability for having a germline or somatic event at the site, given the analyzed tumor type. Finally, previously rejected SSNVs or somatic indels that are more likely to be somatic are recovered.