aiqbench takeMachine Learning

This study demonstrates that the choice of genetic distance encoding—specifically JC69 versus BLAST-derived metrics—fundamentally alters the accuracy of phylogenetic tree reconstruction when using recursive normalized cut methods. The findings highlight that simply optimizing graph-building algorithms is insufficient; the mathematical representation of sequence similarity is a critical variable in evolutionary history tracing.

Sources coveringQuantum Zeitgeist, arXiv

What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Keio University Team Finds Graph Build Impacts Tree Accuracy

Quantum ZeitgeistSeptember 23, 20264 min read

Did a commonly used method for building evolutionary trees previously outperform newer approaches despite relying on less information? Neighbour Joining typically achieved greater accuracy than recursive Normalized Cut reconstructions using standard genetic distance measures like WAG and JC69. However, this new work reveals recursive Ncut surpasses it when utilising BLAST-derived logarithmic distances under certain conditions, highlighting how data encoding impacts results.

Phylogenetic reconstruction accuracy depends upon representations of sequence similarity used. Converting pairwise sequence data into a graph substantially impacts tree accuracy during recursive normalized cut analysis, demonstrating improved reconstructions through informed choices about representing molecular evolution. Specifically, JC69-based local affinities proved key even with increasing divergence whilst other approaches declined in performance. Analysing how genetic data is represented fundamentally impacts reconstructing evolutionary histories and simply…

Excerpt shown under the publisher's RSS terms · 65 words · full text and figures are on the publisher's site

Read at Quantum Zeitgeist ↗Ask about this article