Back

Strand Orientation Bias Detector (SOBDetector) to remove FFPE sequencing artifacts

Diossy, M.; Sztupinszki, Z.; Krzystanek, M.; Eklund, A. C.; Csabai, I.; Pedersen, A. G.; Szallasi, Z.

2019-11-28 bioinformatics
10.1101/386417 bioRxiv
Show abstract

Formalin-fixed paraffin-embedded (FFPE) tissue, the most common tissue specimen stored in clinical practice, presents challenges in the analysis due to formalin-induced artifacts. Here we present SOBDetector, that is designed to remove these artifacts from a list of variants, which were extracted from next-generation sequencing data, using the posterior predictive distribution of a Bayesian logistic regression model trained on TCGA whole exomes. Based on a concordance analysis, SOBDetector outperforms the currently publicly available FFPE-filtration techniques (Accuracy: 0.9 {+/-} 0.015, AUC = 0.96). SOBDetector is implemented in Java 1.8 and is freely available online.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.