# Y matrices
#
gexp_obj = GeneExpression()
gexp = gexp_obj.filter()
LOG.info(f"Gexp: {gexp.shape}")
prot_obj = Proteomics()
prot = prot_obj.filter()
prot = prot[prot.count(1) > 300]
LOG.info(f"Prot: {prot.shape}")
# X matrices
#
crispr_obj = CRISPR()
crispr = crispr_obj.filter(dtype="merged")
LOG.info(f"CRISPR: {crispr.shape}")
drespo_obj = DrugResponse()
drespo = drespo_obj.filter()
drespo = drespo[drespo.count(1) > 300]
drespo = drespo[["+" not in i for i in drespo.index]]
drespo.index = [";".join(map(str, i)) for i in drespo.index]
dtargets = drespo_obj.drugresponse.groupby(
["drug_id", "drug_name", "dataset"])["putative_gene_target"].first()
dtargets.index = [";".join(map(str, i)) for i in dtargets.index]
LOG.info(f"Drug: {drespo.shape}")
# Covariates
#
drug_obj = DrugResponse()
# Samples
#
samples = set.intersection(set(prot_obj.get_data()))
LOG.info(f"Samples: {len(samples)}")
# Filter data-sets
#
prot = prot_obj.filter(subset=samples)
LOG.info(f"Proteomics: {prot.shape}")
gexp = gexp_obj.filter(subset=samples)
LOG.info(f"Transcriptomics: {gexp.shape}")
crispr = crispr_obj.filter(subset=samples, dtype="merged")
LOG.info(f"CRISPR: {crispr.shape}")
drespo = drug_obj.filter(subset=samples)
drespo = drespo.set_index(
pd.Series([";".join(map(str, i)) for i in drespo.index]))
drespo_maxc = drug_obj.maxconcentration.copy()
drespo_maxc.index = [
";".join(map(str, i)) for i in drug_obj.maxconcentration.index
]
drespo_maxc = drespo_maxc.reindex(drespo.index)
LOG.info(f"Drug response: {drespo.shape}")
cn = cn_obj.filter(subset=samples.intersection(prot_obj.ss.index))
cn = np.log2(cn.divide(prot_obj.ss.loc[cn.columns, "ploidy"]) + 1)
elif gene in ESS_GENES:
return "essential"
elif gene in NESS_GENES:
return "non-essential"
else:
return "unclassified"
if __name__ == "__main__":
# Project score
#
cscore_obj = CRISPR()
cscore = cscore_obj.filter(dtype="merged")
cscore_ht29 = cscore["SIDM00136"]
# Samplesheet
#
lib_name = "2gCRISPR_Pilot_library_v2.0.0.xlsx"
lib_ss = pd.read_excel(f"{DPATH}/gi_samplesheet.xlsx")
lib_ss = lib_ss.query(f"library == '{lib_name}'")
lib = read_gi_library(lib_name)
lib["sgRNA1_class"] = [classify_gene(g, c) for g, c in lib[["sgRNA1_Approved_Symbol", "sgRNA1_Chr"]].values]
lib["sgRNA2_class"] = [classify_gene(g, c) for g, c in lib[["sgRNA2_Approved_Symbol", "sgRNA2_Chr"]].values]
lib["vector_class"] = lib["sgRNA1_class"] + " + " + lib["sgRNA2_class"]
samples = list(set(lib_ss["name"]))
TPATH = pkg_resources.resource_filename("tables", "/")
if __name__ == "__main__":
# Data-sets
#
prot_obj = Proteomics()
prot = prot_obj.filter()
LOG.info(f"Proteomics: {prot.shape}")
gexp_obj = GeneExpression()
gexp = gexp_obj.filter(subset=list(prot))
LOG.info(f"Transcriptomics: {gexp.shape}")
crispr_obj = CRISPR()
crispr = crispr_obj.filter(subset=list(prot))
LOG.info(f"CRISPR: {crispr.shape}")
drespo_obj = DrugResponse()
drespo = drespo_obj.filter()
drespo.index = [";".join(map(str, i)) for i in drespo.index]
dmax = drespo_obj.drugresponse.groupby(["drug_id", "drug_name", "dataset"])[
"max_screening_conc"
].first()
dmax = (dmax * 0.5).pipe(np.log)
dmax.index = [";".join(map(str, i)) for i in dmax.index]
dtargets = drespo_obj.drugresponse.groupby(["drug_id", "drug_name", "dataset"])[
"putative_gene_target"