DiscoverY: a classifier for identifying Y chromosome sequences in male assemblies

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• The Y chromosome is crucial for male sex determination and fertility but remains understudied due to its complex genetic structure.
• Existing methods for isolating Y-specific DNA (contigs) from genome assemblies have limitations, including susceptibility to false positives or reliance on high-depth female sequencing data.

Purpose of the Study:

• To develop a robust and versatile method for isolating Y-chromosomal contigs from whole-genome assemblies.
• To overcome the limitations of current Y-chromosome identification strategies, particularly in scenarios with incomplete or low-quality female reference genomes.

Main Methods:

• Developed DiscoverY, a k-mer based computational method.
• Integrated proportion sharing with female genomes and depth of coverage from male reads for contig classification.
• Evaluated performance across human and gorilla genomes using various sequencing platforms (Illumina, 10X, PacBio).

Main Results:

• DiscoverY demonstrates high precision and recall in isolating Y-specific contigs when high-quality male and female data are available.
• The method shows robust performance even when utilizing draft female references or raw female sequencing reads.
• DiscoverY outperforms existing Y-chromosome identification methods under various data quality conditions.

Conclusions:

• DiscoverY provides an effective solution for isolating Y-specific contigs from genome assemblies.
• Regions of the Y chromosome homologous to the X chromosome continue to present challenges for accurate detection.