Harmonise JSON output between Staph and TB

[iqbal-lab]

Spoke to Simon. Simplest to make final release changes to Mykrobe UI if have same JSON format for both.

[Phelimb]

OK, good to know. Will work on it.

[Phelimb]

Can I propose we kill stdout altogether while we do this and make json mandatory? I haven't been keeping it up to date and I fear it's a bit of a mess.

[iqbal-lab]

Agreed. And remove timestamps?

[Phelimb]

I think timestamps are stdout only anyway so removing the option will also remove timestamps.

[Phelimb]

Shall we include a version string in the JSON ?

[Phelimb]

Updated JSONs.

Staph

{
    "expected_depth": "77",
    "mean_read_length": "48",
    "phylogenetics": {
        "complex": {
            "Staphylococcus": "82"
        },
        "species": {
            "S. aureus": "82"
        },
        "lineage": {
            "N/A": "-1"
        }
    },
    "susceptibility": {
        "Gentamicin": "S",
        "Penicillin": "R",
        "Methicillin": "R",
        "Trimethoprim": "S",
        "Erythromycin": "R",
        "FusidicAcid": "S",
        "Ciprofloxacin": "R",
        "Rifampicin": "S",
        "Tetracycline": "S",
        "Vancomycin": "S",
        "Mupirocin": "S",
        "Clindamycin": "R(inducible)"
    },
    "called_variants": {
        "gyrA_S84L": {
            "R_per_cov": "100",
            "S_per_cov": "0",
            "R_median_cov": "101",
            "S_median_cov": "0",
            "conf": "1092",
            "induced_resistance": "Ciprofloxacin"
        }
    },
    "called_genes": {
        "blaZ": {
            "per_cov": "74",
            "median_cov": "82",
            "conf": "42298",
            "induced_resistance": "Penicillin"
        },
        "ermC": {
            "per_cov": "99",
            "median_cov": "834",
            "conf": "2496",
            "induced_resistance": "Erythromycin"
        },
        "mecA": {
            "per_cov": "99",
            "median_cov": "96",
            "conf": "285",
            "induced_resistance": "Methicillin"
        }
    },
    "virulence_toxins": {
        "PVL": "negative"
    }
}

TB

{
    "expected_depth": "58",
    "mean_read_length": "91",
    "phylogenetics": {
        "complex": {
            "MTBC": "100"
        },
        "species": {
            "M. tuberculosis": "63"
        },
        "lineage": {
            "Delhi/Central Asia": "79"
        }
    },
    "susceptibility": {
        "Isoniazid": "S",
        "Rifampicin": "S",
        "Ethambutol": "S",
        "Pyrazinamide": "S",
        "Quinolones": "S",
        "Streptomycin": "S",
        "Amikacin": "S",
        "Capreomycin": "S",
        "Kanamycin": "S"
    },
    "called_variants": {},
    "called_genes": {},
    "virulence_toxins": {}
}

[iqbal-lab]

Version string is good. Commit would be ideal, but hard to organise

[iqbal-lab]

Don't understand what the bug was in calling non-aurues staphs?

[iqbal-lab]

Whats the complex for Staph? If we want to mirror what we do in TB, where the complexes are the two groups we care about, then one complex is S. aureus and the other is coag-neg, right?

[Phelimb]

Bug was that we never called them because we masked them all out rather than masking out aureus. Fixed now.

I can change staph to say aureus or coag neg. I think from a code level though it's not the exact equivalent since we look for catalase to call it as Staph and then have species panels.

[Phelimb]

Sorry, that bug commit is confusing. The bug fix is only in src/predictor/staph/species.c a tiny change. The rest is removing stdout which I should have done in another commit.

[iqbal-lab]

I see. So the question is - what is complex for. My proposal is we use it to decide if we should print resistance predictions. For Staph you print them if there is any aureus, and for TB you do so if there is any MTBC. What do you think?

[Phelimb]

Yes, sounds good. In the case of pure aureus it just seemed strange to me to have complex : auerus, species : aureus.

[iqbal-lab]

yes, it is weird. We could change it from Complex to In-group and Out-group?

[iqbal-lab]

Oh no, needs to change from Complex to one word, not two options!

[iqbal-lab]

SpeciesGroup?

[Phelimb]

Or Clade?

[iqbal-lab]

I basically don't know properly what clade means. I think clade corresponds to a subtree. We want a word that signifies we have split our "local phylogeny" (eg Staphylococci) into two (aureus and not-aureus). How about Partition?

On 12 February 2015 at 22:51, Phelim Bradley notifications@github.com wrote:

Or Clade?

— Reply to this email directly or view it on GitHub https://github.com/iqbal-lab/myKrobe-predictor/issues/37#issuecomment-74170969 .

[Phelimb]

To be honest, we don't really need it at all for Staph. We didn't have it before and the only reason I added it was to be consistent with TB.

[Phelimb]

So whatever word would need to work for MTBC and NTM too otherwise there's not much point .

[iqbal-lab]

Yes, true. Maybe just Group is safe (if boring). We could leave it empty for Staph I suppose, but I thin Aureus versus CoagNeg works

[iqbal-lab]

This is not worth any more of our time.

[iqbal-lab]

Phylogroup?

[Phelimb]

:+1:

[Phelimb]

Tested on staph species with no change so merged to master.