Walk finds allele at 600x that genotype misses

[ronald-jaepel]

Walk finds ptsI-1 with"min_non_zero_depth": 4, and "median_depth": 26, The reported sequence mapps to ptsI-1 when mapped against the reference panel with bwa mem. DNA_sequence_identity 92% & Protein_identity 97%

But Genotype says:

            "ptsI": [
                {
                    "info": {
                        "copy_number": 0.0, 
                        "contamination_depths": [], 
                        "coverage": {
                            "percent_coverage": 22.45, 
                            "median_depth": 0.0, 
                            "min_non_zero_depth": 15.0
                        }, 
                        "expected_depths": [
                            594.5
                        ]
                    }, 
                    "_cls": "Call.SequenceCall", 
                    "genotype": [
                        0, 
                        0
                    ], 
                    "genotype_likelihoods": [
                        -0.5945, 
                        -99999999.0, 
                        -99999999.0
                    ]
                }
            ], 

Sample is in /well/iqbal/people/ronaldj/NIS/fastqs/SRR1970018/ Cortex binary in /well/iqbal/people/ronaldj/NIS/cortex/SRR1970018/ Atlas_Output is in /well/iqbal/people/ronaldj/NIS/jsons/SRR1970018/log_atlas.out

This happens in 4717 of 4835 Salmonella samples

[ronald-jaepel]

Luckily for me, ptsI isn't relevant for me because it has no resistances linked to it. Only R with mutations inactivating gene/downregulating expression. So no hurry. But it's an interesting bug nonetheless.