UR tag in header of BAM file

[elbayaliyev]

Dear Medaka community,

We are using MEdaka 2.1.1.

Command to run Medaka.
Running medaka tr with options: /pl/active/dashnowlab/projects/TR-benchmarking/work_elbay/conda/medaka-2.1.1-affc80a5a289ab61ae9dd7e3044bd223/bin/medaka tandem --sample_name OMLR22_003.haplotagged OMLR22_003.haplotagged.cram /pl/active/dashnowlab/data/ref-genomes/human_GRCh38_no_alt_analysis_set.fasta /pl/active/dashnowlab/projects/TR-benchmarking/catalogs/test_adotto.bed male

We have a bam file that have UR tag for contigs which shows reference file.

We are getting an error with Medaka that points to the file listed in the UR tag. We no longer have this path because the sample was aligned elsewhere. Does this mean Medaka is ignoring the reference file provided as input? As you see we providing /pl/active/dashnowlab/data/ref-genomes/human_GRCh38_no_alt_analysis_set.fasta as a input not the one in header file of @bam.

Error:
Running medaka tr with options: /pl/active/dashnowlab/projects/TR-benchmarking/work_elbay/conda/medaka-2.1.1-affc80a5a289ab61ae9dd7e3044bd223/bin/medaka tandem --sample_name OMLR22_003.haplotagged OMLR22_003.haplotagged.cram /pl/active/dashnowlab/data/ref-genomes/human_GRCh38_no_alt_analysis_set.fasta /pl/active/dashnowlab/projects/TR-benchmarking/catalogs/test_adotto.bed male OMLR22_003.haplotagged.medaka.vcf
[E::fai_build3_core] Failed to open the file /pl/active/dashnowlab/work/ealiyev/Sandbox/nextflow-test/OMLR22_003/work/0a/cf9b729dece2845203294ef320638b/human_GRCh38_no_alt_analysis_set.fasta : No such file or directory
[E::refs_load_fai] Failed to open reference file '/pl/active/dashnowlab/work/ealiyev/Sandbox/nextflow-test/OMLR22_003/work/0a/cf9b729dece2845203294ef320638b/human_GRCh38_no_alt_analysis_set.fasta'
[W::cram_get_ref] Failed to populate reference "chr20"
[W::cram_get_ref] See https://www.htslib.org/doc/reference_seqs.html for further suggestions
[E::cram_next_slice] Failure to decode slice
[15:06:50 - ConsensusGenerator] tr_chr20_155669_155747_pad_155659_155757_fwd_hap0_phased-set0_ploidy2: Retrieved too few reads (0 < 3)
[E::fai_build3_core] Failed to open the file /pl/active/dashnowlab/work/ealiyev/Sandbox/nextflow-test/OMLR22_003/work/0a/cf9b729dece2845203294ef320638b/human_GRCh38_no_alt_analysis_set.fasta : No such file or directory

Thanks,
Elbay

[hdashnow]

Hi Medaka team,

We'd like to run Medaka tandem on a large cohort. While we can manage this issue to some extent by reheadering the file, this won't scale to the cohort size we need to run this on. Are there any plans for a fix, or can you recommend another workaround?

Warm regards,
Harriet

[snurk]

Dear @elbayaliyev and @hdashnow, thank you for your interest in medaka tandem and sorry the issue is taking a while to resolve. We will be looking into improved CRAM support shortly and will try to release an update by the end of February.