Skip to content

Region targets #21

Description

@esteinig
  • add target regions logic using minimap2-rs
  • add support for specifying targets in config - deserialize into struct
  • deprecate sam mapping logic - not using dorado fork
  • test previous human chromosome targets that change to logic still works
  • add file input for lists of targets
  • renamed target struct "name" field to "reference" in config.rs and added optional "name" field for target region specification to better keep track of which e.g. gene is targeted
  • add better log message to experiment input
  • test target file input
  • test on a single chromosome run
  • add evaluation and coverage plots for regions

Target file format (tab-delimited), no header, can be mixed formats:

  1. Single field with reference name (sequence name present in index) (no target coordinates)
  2. Target coordinates on reference sequence, fields:
    • reference (sequence name present in index)
    • start (1-based, closed)
    • end (1-based, closed)
    • name (assigned name of target sequence)

When specified in TOML config list (targets) can also be mixed formats:

  1. reference (no target coordinates)
  2. {reference}::{start}::{end}::{name} (1-based, closed)

Region alignment logic must consider:

  • Start inside target range: alignment start >= target range start and alignment start <= target range end
  • End inside target range: alignment end >= target range start and alignment end <= target range end
  • Spans target range: alignment start <= target range start and alignment end >= target range end

Modified evaluation scripts:

  • separate alignment with minimap2 and paf output
  • consideration of secondary alignment logic as in decision logic for unblocking
  • rework to always consider target file with optional regions - specifies a target with start=0 and end=0 to be considered mapped if the target reference matches (naive / broad cases for targeted sequencing) as previously implemented without coordinates - in the unblock decision logic this is considered as optional start end values of target specifications in config

Metadata

Metadata

Assignees

Labels

No labels
No labels

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions