samblaster not found

[CeciliaDeng]

Description of the bug

Hi @GallVp , AssemblyQC successfully completed assessing an assembly with HiC data. When I check the HiC QC report, it mentioned [forward/reverse] Hi-C reads not found, as well as 'samblaster command not found'.

Is somehow samblaster not loaded?
Thank you.

Command used and terminal output

sbatch pfr_assemblyqc
Submitted batch job 7756409

Relevant files

No response

System information

No response

[GallVp]

Thank you for submitting an issue on this. I have looked at the report for the pipeline test data and it also has the same items missing.

I think this information is missing because we don't provide the path to the raw reads to the HICQC module. I'll investigate and see if this can be fixed.

Nonetheless, if Subjective Hi-C library judgment is given along with other statistics that means that HICQC had enough information to reach a judgement on the quality of the library. See the following link for interpreting the judgement: https://phasegenomics.github.io/2019/09/19/hic-alignment-and-qc.html

[GallVp]

Versions

assemblyqc version: 2.1.1
hic_qc version: https://github.com/phasegenomics/hic_qc/tree/6881c3390fd4afb85009a52918b4d068100c58b4
hic_qc container: https://github.com/GallVp/docker-images/blob/76f2da6af94c2899849ac13fd82decc43401bed5/hic_qc/6881c33_ps/Dockerfile#L9

Problems

• Forward and reverse reads file names are not parsed from the BAM header by hic_qc if the fasta file names don't end with _r1.fq.gz and _r2.fq.gz. This behaviour is codified here: https://github.com/phasegenomics/hic_qc/blob/6881c3390fd4afb85009a52918b4d068100c58b4/hic_qc.py#L320C79-L320C85

• hic_qc assumes that a reference file is provided to the BWA mem command and it uses this assumption to parse the name of the reference file from the BAM header. Whereas, assemblyqc provides the BWA index which is ignored by hic_qc parser. This is codified here: https://github.com/phasegenomics/hic_qc/blob/6881c3390fd4afb85009a52918b4d068100c58b4/hic_qc.py#L319

Solution

• Improve the hic_qc header parser and create a PR on the hic_qc repo: https://github.com/phasegenomics/hic_qc.
• Create a new container for hic_qc or submit to Bioconda.
• Update assemblyqc HICQC module:

  assemblyqc/modules/local/hicqc.nf

  Line 5 in cec2728

  container "docker.io/gallvp/hic_qc:6881c33_ps"

[GallVp]

The results of the report are correct. Only information regarding reference and FastQ file names is missing. The solution requires changes in the tool repo. Therefore, I am placing this issue on backlog and deprioritising it.