Skip to content

Latest commit

 

History

History
114 lines (83 loc) · 4.79 KB

TODO.md

File metadata and controls

114 lines (83 loc) · 4.79 KB
Raw

TODO

  • Read from stdin, useful for quick look ups like intersectBed -a x.vcf -b y.bed | ASCIIGenome -. In practice, this would simply read from stdin and write to a temp file which is then sorted, compressed and indexed. Ok for small files, fewer then 1/2M records (?), inpractical otherwise. You need to guess the format of the input. Unless you enforce syntax like ASCIIGenome -.vcf, which is awkward. ON HOLD - see post on jline-users.

  • Display soft clipped bases.

FIXME

  • TrackMethylation: method getRecordsAsStrings() hard-coded returns empty list. Implement it properly or extend TrackMethylation to TrackReads

Features

  • 17-08-2023: Allow whitespace (in addition to tab) to be a column separator in bed/bedgraph. Do it in MakeTabixIndex.sortByChromThenPos (if a file is sorted and tabix-indexed it is also tab-separated since tabix only allows tab)

  • 01-08-2023: Add a moveTracks command to reorder tracks. You probably need a TrackSet.moveTracks(cmdTokens) method similar to TrackSet.orderTracks(...). CLI:

moveTracks FOO BAR --after SPAM
moveTracks FOO BAR --before SPAM
moveTracks --before SPAM # Error: Provide tracks to move

moveTracks FOO BAR # Swap order of FOO and BAR or throw error `provide --before/--after track`

moveTracks FOO --after NONEXISTS # Error: No track matching NONEXISTS
moveTracks FOO BAR --after FOO # Error: Cannot move a track before or after itself
moveTracks FOO FOO --after BAR # Make sure FOO is not duplicated

# Should the order of tracks to move matter? 
# Should these command return the same order?
moveTracks FOO SPAM --after BAR
moveTracks SPAM FOO --after BAR
``

* 01-07-2023: Allow the first line in bed/bedgraph/gtf to fail under assumption it is the header

* Use `samtools depth` to compute coverage in `TrackPileup` over large regions. Use of `samtools depth` is avoided when filters except those in `samtools` command 
are enabled.

* Shade base quality to optionally shade a range of qualities with different shades of grey. 
  `shadeBaseQuality -below_baseq x | -range` where `-below_baseq` shades below bseq *x* as it is now.
  `-range` sets a level of grey according to quality.

* Converter for ensembl to ucsc chromosome names. This functionality could be included
  in `addTracks` by adding options `-ens2ucsc` and `-ucsc2ens`. E.g. `addTracks -ens2ucsc genes.gtf http://foo/bar/data.bigwig`.
  The option `-ens2ucsc|-ucsc2ens` should iterate through the target file and make an edited copy 
  where each line is converted as appropriate. Then load the edited copy by re-sending it to `addTracks`. 
  For the conversion use the conversion files also used by IGV, put them in resources.
  There are quite a few limitations: Large files, especially bams, will take ages; TDF and bigwig
  cannot easily converted, especially TDF.
  Option 2: Add a field in the `Track` class to indicate that the input file needs to
  be converted. When you fetch a region send to the track the command with the chromosome
  name changed. 

* Deprecate `-fa` option? If the list of files includes a fasta file use that as reference

* Time out for `updateCheck()`. Something like:

long now= System.currentTimeInMiilis(); while(System.currentTimeInMiilis() < (now + 10000)){ updateCheck(); }


* Allow the user to set the `tmp` directory.

* Option to log transform quantitative data.

* Sampling feature: Command to display only a random sample of features in current window. 
  Useful to see the density of features in large intervals with lots of features. E.g.

sample INT|percent [track_regex = .*] ...


This cmd should operate on the feature set after grep selection. Probably `TrackIntervalFeature.getFeaturesInInterval()` is where
to work for this.

* `export` command to write to file the data in the current interval? API:

export [track_regex = .*] > [outdir/]


Each track selected by regex will be written to outdir as `chrom_start-end.track_tag.ext` with appropriate `ext`ension.
Coverage and wiggle files will be bedgraph with the score column taken from the Track. Annotation files will bed or gtf
with data taken from the raw. Reference sequence if available will be in fasta format.

* Command to "go to other end" of feature? 

* Add a `nucPrint` command to print nucleotide counts at each position, a bit like pysamstats. API:
`nucPrint INT [INT] [stranded] [pct]`

Refactor
--------

* Populate the enum class `Command` and use `Command.cmdName` in place of the hardcoded command names.

* Each command in CommandHelp should be a class on its own extending CommandHelp. The super class command help
should implement a `validate()` method which is customized to each sub class. `.validate()` should check e.g.
that the command exists, regex are fine, number of args is ok, etc.