Gene suggestions for 'Clock' yields Per2

[pkerpedjiev]

(hg-server) ===========================================================================================================
[peter@mbi-cw-l10381 higlass-server] [master]$ python
Python 2.7.11 (default, Nov 17 2016, 17:37:33)
[GCC 4.2.1 Compatible Apple LLVM 8.0.0 (clang-800.0.42.1)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import tilesets.suggestions as ts
>>> ts.get_gene_suggestions('data/gene-annotations-mm9.db', 'Clock')
[{'chr': u'chr1', 'score': 394.0, 'geneName': u'Per2', 'txEnd': 93355905, 'txStart': 93312558}, {'chr': u'chr5', 'score': 308.0, 'geneName': u'Clock', 'txEnd': 76733817, 'txStart': 76638892}, {'chr': u'chr11', 'score': 283.0, 'geneName': u'Per1', 'txEnd': 68923459, 'txStart': 68912457}, {'chr': u'chr4', 'score': 69.0, 'geneName': u'Per3', 'txEnd': 150418774, 'txStart': 150377763}, {'chr': u'chr10', 'score': 41.0, 'geneName': u'Timeless', 'txEnd': 127689997, 'txStart': 127669118}, {'chr': u'chr12', 'score': 19.0, 'geneName': u'Cipc', 'txEnd': 88306316, 'txStart': 88287990}]

[mccalluc]

Nils reported a similar issue, searching for "pam". Searching for "perm" or "tan" causes the same kind of problem. higlass/higlass#87

[pkerpedjiev]

This is because genes are ranked by importance. As long as the query text is included in the gene or its description, the results are ranked by the importance of the gene.

The fix would be to give extra weight to search terms which match the gene name exactly.