Lucas Robidou

This paper revisits the concept of density in k-mer-based sequence analysis, introduces multiminimizers as a novel technique, and proposes deduplicated density as a new metric.

The paper introduces Super Bloom, a Bloom filter variant for streaming k-mer queries, which improves cache efficiency and reduces false positives.

The paper introduces hyper-k-mers, a more efficient k-mer representation, and presents KFC, a tool that significantly reduces memory usage and improves speed for k-mer counting.

The paper introduces fimpera, an algorithm that improves counting Approximate Membership Query data structures by reducing false positives and filter size while maintaining precision.

findere reduces false-positive rates in approximate membership queries for k-mers without increasing space or memory usage.