LexicMap 0.4.0

v0.4.0 - 2024-08-15

• New commands:
  • lexicmap utils 2blast: Convert the default search output to blast-style format.
• lexicmap index:
  • Support suffix matching of seeds, now seeds are immune to any single SNP!!!, at the cost of doubled seed data.
  • Better sketching desert filling for highly-repetitive regions.
  • Change the default value of --seed-max-desert from 900 to 200 to increase alignment sensitivity.
  • Mask gap regions (N's).
  • Fix skipping interval regions by further including the last k-1 bases of contigs.
  • Fix a bug in indexing small genomes.
  • Change the default value of -b, --batch-size from 10,000 to 5,000.
  • Improve lexichash data structure.
  • Write and merge seed data in parallel, new flag -J/--seed-data-threads.
  • Improve the log.
• lexicmap search:
  • Fix chaining for highly-repetitive regions.
  • Perform more accurate alignment with WFA.
  • Use buffered reader for seeds file reading.
  • Fix object recycling and reduce memory usage.
  • Fix alignment against genomes with many short contigs.
  • Fix early quit when meeting a sequence shorter than k.
  • Add a new option -J/--max-query-conc to limit the miximum number of concurrent queries,
    with a default valule of 12 instead of the number of CPUs, which reduces the memory usage
    in batch searching.
  • Result format:
    • Cluster alignments of each target sequence.
    • Remove the column seeds.
    • Add columns gaps, cigar, align, which can be reformated with lexicmap utils 2blast.
• lexicmap utils kmers:
  • Fix the progress bar.
  • Fix a bug where some masks do not have any k-mer.
  • Add a new column prefix to show the length of common prefix between the seed and the probe.
  • Add a new column reversed to indicate if the k-mer is reversed for suffix matching.
• lexicmap utils masks:
  • Add the support of only outputting a specific mask.
• lexicmap utils seed-pos:
  • New columns: sseqid and pos_seq.
  • More accurate seed distance.
  • Add histograms of numbers of seed in sliding windows.
• lexicmap utils subseq:
  • Fix a bug when the given end position is larger than the sequence length.
  • Add the strand ("+" or "-") in the sequence header.