cpToGrp[brd] = drugLabels['class'][ibrd]
inameDict = {}
for ibrd, brd in enumerate(drugLabels['pert_id']):
inameDict[brd] = drugLabels.ix[ibrd]['pert_iname']
### cell line match mode
wkdir = '/xchip/cogs/sig_tools/sig_summly/pcl/dw_sep_24'
if not os.path.exists(wkdir):
os.mkdir(wkdir)
reload(pcla)
metric = 'wtcs'
po = pcla.PCLA(grpToCp,
metric,
wkdir,
summly_out_prefix='summly_out',
pairwise_prefix='pairwise_matrices',
cell_match_mode=True,
row_space='lm')
po.get_sig_ids()
# po.run_summly(rerun_mode=False)
# summPath = po.out + '/summly_out/sep11'
summPath = '/xchip/cogs/projects/connectivity/summly/matched/src'
po.make_summly_path_dict(summPath)
# po.run_summly(rerun_mode=True)
# po.make_summly_path_dict(summPath_nMtch)
po.get_inames()
# po.test_groups(make_heatmaps=True,
# group_size_min=3,
# sum_score_metric='sum_score_4',
# rankpt_metric='mean_rankpt_4')
ttd_cp_dict[groupName + '-' + groupCat] = cpLst
inameDict = {}
for x in drugLabels.iterrows():
brds = x[1]['pert_ids_merged']
iname = x[1]['Drug_name']
for brd in brds:
inameDict[brd] = iname
reload(pcla)
gp_type = 'KD'
metric = 'wtcs'
po = pcla.PCLA(ttd_cp_dict,
metric,
wkdir,
summly_out_prefix='summly_out',
pairwise_prefix='pairwise_matrices_by_Category',
cell_match_mode=True,
row_space='lm')
po.get_sig_ids()
# po.run_summly(rerun_mode=False)
# summPath = po.out + '/summly_out/sep11'
summPath = '/xchip/cogs/data/rnwork/batch_summly/summly_lm50'
po.make_summly_path_dict(summPath)
# po.run_summly(rerun_mode=True)
# po.make_summly_path_dict(summPath)
po.inameDict = inameDict
po.test_groups(make_heatmaps=False,
group_size_min=15,
sum_score_metric='sum_score_4',
rankpt_metric='mean_rankpt_4')
### load in data for individual groups
llo = ldc.label_loader()
pclDict = llo.load_drugbank_by_gene(group_by_action=False)
#get updated pcl dict (to summly space) using pcla module
### class analysis
wkdir = '/xchip/cogs/projects/pharm_class/TTd_Oct29'
if not os.path.exists(wkdir):
os.mkdir(wkdir)
reload(pcla)
metric = 'wtcs'
po = pcla.PCLA(pclDict,
metric,
wkdir,
pairwise_prefix='pairwise_matrices',
rankpt_metric='mean_rankpt_4',
sum_score_metric='sum_score_4',
row_space='lm',
cell_match_mode=True)
po.get_inames()
po.load_summly_mtrx()
summDict = po.pclResultDict
dbGenes = summDict.keys()
dbGenes.remove('-666')
#look up baseline expression of these genes
MC = mu.MongoContainer()
ci = MC.gene_info.find({'pr_gene_symbol': {
'$in': dbGenes
}}, {
'is_expressed': True,