findere

findere reduces false positives and accelerates queries on Approximate Membership Query (AMQ) structures such as Cuckoo filters and Bloom filters to improve accuracy and performance for large-scale k-mer and set-membership analyses.

Key Features:

• AMQ compatibility: Operates with Approximate Membership Query (AMQ) structures including Cuckoo filters and Bloom filters.
• False-positive reduction: Dynamically reduces the false-positive rate by two orders of magnitude at query time without altering the original indexing data structure.
• Query speed enhancement: Accelerates query processing by a factor of two.
• No false negatives or extra memory: Ensures no false negatives while not requiring additional memory overhead.
• User-defined trade-offs: Allows adjusting either the false-positive rate or the space required to represent a set according to user constraints.

Scientific Applications:

• k-mer indexing: Improves the accuracy and efficiency of k-mer presence/absence indexing in large sequencing-derived datasets.
• Genomic and large-scale biological data processing: Enhances set-membership queries for genomic analyses and other high-throughput biological data workflows.

Methodology:

Applies a query-time filtering strategy that dynamically reduces AMQ false positives without modifying the underlying AMQ index or allocating additional memory.