Solve GI phase out and FutureDeprecation warning

This was (presumably) the last piece of code to change to make the
whole pipeline work without GIs after their phase out. Further,
`sort_values` is used instead of `order` from `pandas`, thus squashing
the warning that appeared.

Fix #16 and fix #17.

virmet/covplot.py

@@ -10,17 +10,17 @@
 
 
 def best_species(orgs_file, organism):
-    orgs_list = pd.read_csv(orgs_file, sep=',', header=0)
+    orgs_list = pd.read_csv(orgs_file, sep='\t', header=0)
     # assert decreasing sorted
-    diff = orgs_list['0'] - orgs_list['0'].shift(1)
+    diff = orgs_list['reads'] - orgs_list['reads'].shift(1)
     assert (diff > 0).sum() == 0, diff
     # criterion is "startswith"
     # criterion = orgs_list['sscinames'].map(lambda x: x.startswith(organism))
     criterion = orgs_list.iloc[:, 0].str.startswith(organism).fillna(False)
     matching_orgs = orgs_list[criterion]
     # organism matching that given on command line with most reads is the first
     # W.O. this assumes descending order of reads
-    return str(matching_orgs.iloc[0].sscinames)
+    return str(matching_orgs.iloc[0].organism)
 
 
 def main(args):
@@ -37,10 +37,9 @@ def main(args):
     blast_file = os.path.join(outdir, 'unique.tsv.gz')
     unique = pd.read_csv(blast_file, sep='\t', header=0, compression='gzip')
     matching_reads = unique[unique['sscinames'] == best_spec]
-    best_seqids = matching_reads.groupby('sseqid').size().order(ascending=False)
+    best_seqids = matching_reads.groupby('sseqid').size().sort_values(ascending=False)
 
-    # TODO: upgrade for NCBI outphase of GI
-    gi, dsc, acc = str(best_seqids.index.tolist()[0]).split('|')[1:4]
+    dsc, acc = str(best_seqids.index.tolist()[0]).split('|')[:2]
 
     # copy single genome, index, align viral_reads
     os.chdir(outdir)