wdecoster/cramino

A *fast* tool for BAM/CRAM quality evaluation, intended for long reads

cramino's Issues

Mean identity calculated as more than 100%

I'm seeing several cases where mean identity is calculated as over 100%. For example:

Yield [Gb]	39.86
Mean coverage	12.86
N50	21602
Median length	1600.00
Mean length	5308
Median identity	97.98
Mean identity	103.01

I wouldn't expect this to be possible, correct?

Support older glibc

Identities and splicing

How does the calculation of the identities take splicing into account?

thread 'main' panicked at 'Unexpected type of Aux I32(8) error

Hi,

I just wanted to try out Cramino and I installed it using conda. However, on testing it on one of the bam files produced using pbmm2, I am getting this error-

thread 'main' panicked at 'Unexpected type of Aux I32(8)', src/extract_from_bam.rs:164:18 note: run with 'RUST_BACKTRACE=1' environment variable to display a backtrace

This is the cramino command I am using-

time cramino -t 8 --hist --karyotype second_hg38.movie.bam

Any workaround? Is it a bug in the code by any chance? I haven't installed Rust yet, but I guess that it should be installed separately?

Regards,
Prasun

PS-Just to add, I installed Rust (v1.70.0), but the error still persists. I am using cramino v0.9.7 which I installed using conda

--spliced not supported? (Question)

Is the --spliced supported?, I tried to use but I keep getting this error

error: Found argument '--spliced' which wasn't expected, or isn't valid in this context
If you tried to supply --spliced as a value rather than a flag, use -- --spliced

Plus, it no longer shows in the command line help (-h) message.

Feature request: filter on rlen/qlen

Filter reads on a certain rlen and/or qlen (cli options).

Does cramino support unmapped bam?

It works with the cram file I have but crashes when I give it an unmapped bam:

~/utils/cramino -t 4 ABC.ubam --arrow ABC.unmapped.feather
File name ABC.ubam
[2023-07-21T04:19:31Z ERROR cramino] Not enough reads to calculate metrics!
thread 'main' panicked at 'explicit panic', src/main.rs:155:9
note: run with RUST_BACKTRACE=1 environment variable to display a backtrace

By the way, is it possible to also output number of mapped bases and unmapped bases for aligned cram/bam?

Improve histograms: use integers for read length and less decimals for identities

Currently using https://github.com/fgadaleta/histo_fp

More dynamically handle number of bins for read length histogram

Currently hardcoded at 100 bins, but this is problematic with off-label use for e.g. short reads, amplicons or a tiny tiny cram/bam.
Number of bins should be min(maxlen-minlen, 100)

Optionally output as json

Optionally output the metrics as a (pretty printed) json