Back

Random forest classifiers trained on simulated data enable accurate short read-based genotyping of structural variants in the alpha globin region at Chr16p13.3

Hansen, N. F.; Wang, X.; Tegegn, M. B.; Liu, Z.; Gouveia, M. H.; Hill, G.; Lin, J. C.; Okulosubo, T.; Shriner, D.; Thein, S. L.; Mullikin, J. C.

2023-11-27 bioinformatics
10.1101/2023.11.27.568683 bioRxiv
Show abstract

In regions where reads dont align well to a reference, it is generally difficult to characterize structural variation using short read sequencing. Here, we utilize machine learning classifiers and short sequence reads to genotype structural variants in the alpha globin locus on chromosome 16, a medically-relevant region that is challenging to genotype in individuals. Using models trained only with simulated data, we accurately genotype two hard-to-distinguish deletions in two separate human cohorts. Furthermore, population allele frequencies produced by our methods across a wide set of ancestries agree more closely with previously-determined frequencies than those obtained using currently available genotyping software.

Matching journals

The top 11 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.