I’m converting the two column tab delimited file as JSON format, together with set of keys and argument sample_id = 'WGNP1000001'.

Here’s my input and expected output format and the code. Appreciate any help.

Python to convert tab delimited file as json with supplied argument and additional keys

Tab delimited input file: WGNP1000001.list.txt

insert_size 447.3
insert_size_std 98.2
pct_properly_paired 97.9
pct_mapped  99.63
yield_bp_q30    1996315
mean_autosome_coverage  0.000644
pct_autosomes_15x   0.000016
mad_autosome_coverage   4

Expected Output without adding double quotes to the values and not adding float to the integer only values eg. mad_autosome_coverage, yield_bp_q30 and pct_autosomes_15x

{
    "sample": {
        "id": "WGNP1000001"
    },
    "wgs_metrics": {
        "insert_size": 447.3,
        "insert_size_std": 98.2,
        "mad_autosome_coverage": 4,
        "mean_autosome_coverage": 0.000644,
        "pct_autosomes_15x": 0.000016,
        "pct_mapped": 99.63,
        "pct_properly_paired": 97.9,
        "yield_bp_q30": 1996315
    }
}

But the following tried code writes (try1) the values with double quotes or (try2) adding the float to the the values of integer only.

Try1 – write the values with double quotes

def raw_data(input_metrics):
    d = {}
    with open(input_metrics) as f:
        for line in f:
            if not line.strip():
                continue
            row = line.split('t')
            key = row[0]
            value_str = row[1]
            d[key] = value_str.replace("n", "")
            #try:
            #    value = float(value_str.strip())
            #except ValueError:
            #    value = value_str.strip()
            #d[key] = value
    return d

Output:

{
    "sample": {
        "id": "NA12878"
    },
    "wgs_metrics": {
        "insert_size": "447.3",
        "insert_size_std": "98.2",
        "mad_autosome_coverage": "4",
        "mean_autosome_coverage": "0.000644",
        "pct_autosomes_15x": "0.000016",
        "pct_mapped": "99.63",
        "pct_properly_paired": "97.9",
        "yield_bp_q30": "1996315"
    }
}

Try2 – adding float to the integer only values (yield_bp_q30, mad_autosome_coverage) and log (pct_autosomes_15x)

    def raw_data(input_metrics):
        d = {}
        with open(input_metrics) as f:
            for line in f:
                if not line.strip():
                    continue
                row = line.split('t')
                key = row[0]
                value_str = row[1]
                try:
                    value = float(value_str.strip())
                except ValueError:
                    value = value_str.strip()
                d[key] = value
        return d

Output:
{
    "sample": {
        "id": "WGNP1000001"
    },
    "wgs_metrics": {
        "insert_size": 447.3,
        "insert_size_std": 98.2,
        "mad_autosome_coverage": 4.0,
        "mean_autosome_coverage": 0.000644,
        "pct_autosomes_15x": 1.6e-05,
        "pct_mapped": 99.63,
        "pct_properly_paired": 97.9,
        "yield_bp_q30": 1996315.0
    }
}

Updated code:

def raw_data(input_metrics):
    d = {}
    with open(input_metrics) as f:
        for line in f:
            if not line.strip():
                continue
            row = line.split('t')
            key = row[0]
            value_str = row[1]
            try:
                value = int(value_str)
            except ValueError:
                value = float(value_str.strip())
            d[key] = value
    return d