Skip to content

Latest commit

 

History

History
26 lines (22 loc) · 1.41 KB

TODO.md

File metadata and controls

26 lines (22 loc) · 1.41 KB

TODO

Issues

  • Flanking sequences are not called correctly
    • Causes spacers to be called as flankers
  • Speed degrades rapidly with the number of patterns identified
    • Probable cause is the singleton finder
  • In some circumstances closely related groups are not getting split apart
  • The final output coverage of the spacer does not always correspond to the number of sources
  • Spacers go missing in the graph from Crass, but they are present if you perform a regular overlap assembly
    • This is not a sequencing error issue as even when searching with mismatches, there is not sign of the spacers
  • Spacer strings only get created during the initial pass of the graph building. However, spacer objects can also be made from the p-graph
    • This means that spacers may exist in the graph, but not in the output sequences. Maybe the cause of the above issue
  • If a slave DR is a perfect palindrome of the master DR the offset will be undefined

Wanted Features

  • Multi-threading the search algorithms
  • Using paired read information in the search algorithms
  • Ability to run Crass on genomes or assembled contigs
    • Output a gff3 formatted file
  • Screen for known eukaryotic microsatellites for datasets that may be host contaminated

Improvements

  • Use read information better when building the graph
  • Allow for multiple DR types to be in a single sequence. Very important if Crass is to work on genomes