summarize_blast_results.py

## summarize_blast_results.py
#!/usr/bin/env python

# Click 'Download > Multiple-file JSON' from NCBI search results page,
# unzip it, run this script in it without any parameters, get the
# markdown formatted table.

import json
import glob

# poor man's whatever:
QUERY    = lambda: hits['BlastOutput2']['report']['results']['search']['query_title'].split('___')[0]
QLEN     = lambda: hits['BlastOutput2']['report']['results']['search']['query_len']
HIT      = lambda: hits['BlastOutput2']['report']['results']['search']['hits'][index]
DESC     = lambda: HIT()['description'][index]
TITLE    = lambda: DESC()['title']
SCINAME  = lambda: '_%s_' % DESC()['sciname']
ACC      = lambda: '[%(desc)s](https://www.ncbi.nlm.nih.gov/protein/%(desc)s)' % {'desc': DESC()['accession']}
HSPS     = lambda: HIT()['hsps'][0]
PCTALIGN = lambda: '%.2f%%' % (HSPS()['align_len']* 100 / QLEN())
PCTID    = lambda: '%.2f%%' % (HSPS()['identity'] * 100 / HSPS()['align_len'])


print('|'.join(['', 'Found in the assembly', 'Best hit on NCBI', 'Percent alignment', 'Percent identity', 'Accession', '']))
print('|'.join(['', ':--', ':--', ':--:', ':--:', ':--:', '']))

# go through every json file in the directory:
for j in glob.glob('*.json'):
    hits = json.load(open(j))

    # skip the poop file
    if 'BlastOutput2' not in hits:
        continue

    # report the best hit:
    index = 0

    # unless the best hit resolves to a multispecies .. if it does, increment
    # index
    while 1:
        if TITLE().find('MULTISPECIES') == -1:
            break

        index += 1

    print('|'.join(['', QUERY(), SCINAME(), PCTALIGN(), PCTID(), ACC(), '']))