Back

Somatic Truth Data from Cell Lineage

Shand, M.; Soto, J.; Lichtenstein, L.; Benjamin, D.; Farjoun, Y.; Brody, Y.; Maruvka, Y. E.; Blainey, P. C.; Banks, E.

2019-10-31 bioinformatics
10.1101/825042 bioRxiv
Show abstract

Existing somatic benchmark datasets for human sequencing data use germline variants, synthetic methods, or expensive validations, none of which are satisfactory for providing a large collection of true somatic variation across a whole genome. Here we propose a dataset of short somatic mutations, that are validated using a known cell lineage. The dataset contains 56,974 (2,687 unique) Single Nucleotide Variations (SNV), 6,370 (316 unique) small Insertions and Deletions (Indels), and 144 (8 unique) Copy Number Variants (CNV) across 98 in silico mixed truth sets with a high confidence region covering 2.7 gigabases per mixture. The data is publicly available for use as a benchmarking dataset for somatic short mutation discovery pipelines.

Matching journals

The top 4 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.