Knime node to fetch molblock and protein name for query and hit fragment.

knime-kripodb-merge.py

import pandas as pd
from kripodb.canned import fragments_by_id

# Requires as input table with columns
# hit_frag_id, query_frag_id and score
# Returns id and molblock of query and hit 
# and protein name of hit.

# Location of fragment database
fragments_db_filename = '/data/kripo/tiny/fragments.sqlite'
# Columns to output
required_cols = ['query_frag_id', 'query_mol', 'hit_frag_id', 'hit_mol', 'score', 'hit_prot_name']


query_table = fragments_by_id(input_table['query_frag_id'], 
                              fragments_db_filename,
                              'query_')

inter_table = pd.merge(input_table, 
                        query_table, 
                        on='query_frag_id')

hits_table = fragments_by_id(input_table['hit_frag_id'], 
                             fragments_db_filename,
                             'hit_')

output_table = pd.merge(inter_table, 
                        hits_table, 
                        on='hit_frag_id')[required_cols]