Skip to content

Align script

rmuntean requested to merge align_script into main

For aligning a sequence there are a few steps:

  • decompress the snapshot '.provision.json.xz' format
  • load a lookup file (key:seq_hash; value:aligned_sequence) with the alignment for all the previously aligned sequences
  • split all the sequences which could not be found in the lookup file in N smaller snapshot modules
  • run NextAlign against each snapshot module in parallel
  • merge together all the aligned sequences (aligned using the lookup table or NextAlign) into a final .provision.json file
  • update the lookup file with the new aligned sequences

Merge request reports