PyDamage: automated ancient damage identification and estimation for contigs in ancient DNA <i>de novo</i> assembly

DNA de novo assembly can be used to reconstruct longer stretches of (contigs), including genes and even genomes, from short sequencing reads. Applying this technique metagenomic data derived archaeological remains, such as paleofeces dental calculus, we investigate past microbiome functional diversity that may absent or underrepresented in the modern gene catalogue. However, compared samples, ancient samples are often burdened with environmental contamination, resulting datasets represent mixtures DNA. The ability rapidly reliably establish authenticity integrity is essential for studies, distinguish between sequences particularly important studies. Characteristic patterns damage, namely fragmentation cytosine deamination (observed C-to-T transitions) typically authenticate sequences, but existing tools inspecting filtering aDNA damage either compute it at read level, which leads high loss lower quality when combination assembly, require manual inspection, impractical assemblies contain tens hundreds thousands contigs. To address these challenges, designed PyDamage, a robust, automated approach estimation authentication assembled aDNA. PyDamage uses likelihood ratio based discriminate truly contigs originating contamination. We test on both simulated paleofeces, demonstrate its automatically identify bearing characteristic Coupled Pydamage opens up new doors explore datasets.