make_GOplots_dataframes.py

##############################################################################
###                     												   ###
### AUTHOR: CONTESSA A. RICCI, PhD 										   ###
### MANUSCRIPT: DYSREGULATION OF MITOCHONDRIA-MEDIATED MATERNAL-FETAL	   ###
###             INTERACTIONS IN HYPERTENSIVE DISORDERS OF PREGNANCY		   ###
###             DOI: (TBD)												   ###
### STUDY PURPOSE: Reanalysis of longitudinal maternal RNAseq data from    ###
###                peripheral blood plasma and endpoint fetal RNAseq data  ###
###                from placenta at delivery. Goal is to understand        ###
###                consequences of mitochondrial gene dysregulation by     ###
###                examining expression patterns of genes the dysregulated ###
###                mitochondrial genes are known to interact with		   ###
### DATA: GEO accession GSE154377 (maternal longitudinal data)			   ###
###       GEO accession GSE114691 (fetal placental)						   ###
###       Accessible via NCBI GEO datasets								   ###
###                     												   ###
### NOTES: All scripts assume all data has been compiled and is in correct ###
###        format. This script utilizes some files that have been compiled ###
###        either manually, in Python, or in R (or a combination thereof). ###
###                     												   ###
##############################################################################

# PURPOSE OF SCRIPT: Create dataframes for GOplot plots in R. Z-scores will be calculated in R.
#
#
# notes:
#		-	will use mtDEG-INTXs expression files generated by find_mtDEG-INTXs.py script:
#												- maternal_del_mtDEG-INTXs_expression.csv
#												- fetal_del_mtDEG-INTXs_expression.csv
#		-	will use GO enrichment files generated using GO enrichment analysis feature at geneontology.org:
#												- maternal_delivery_BPresults.txt (analysis conducted on 12/06/2021)
#												- maternal_delivery_CCresults.txt (analysis conducted on 12/06/2021)
#												- fetal_delivery_BPresults.txt (analysis conducted on 12/06/2021)
#												- fetal_delivery_CCresults.txt.txt (analysis conducted on 12/06/2021)
#				note: Maternal and fetal GO enrichments were determined in separate analyses. GO enrichment
#					  categories (biological process, cellular component) were also determined in separate analyses.
#					  GO enrichment files were generated as follows:
#										- Uniprot IDs outputted using find_mtDEG-INTXs.py script 
#										  (fetal_del_mtDEG-INTXs_UniprotIDs.txt, maternal_del_mtDEG-INTXs_UniprotIDs.txt)
#										  were pasted into "Go Enrichment Analysis" dialog box.
#										- For biological process enrichments "biological process" was selected from the 
#										  dropdown menu immediately below dialog box where Uniprot IDs were pasted. 
#										  Likewise, "cellular component" was selected for cellular component enrichments.
#										- "Homo sapiens" was was specified as species using dropdown menu immediately below
#										  dropdown menu for GO enrichment categories. 
#										- Click the "Launch" button (located on same row as dropdown menu for species
#										  specification) to run analysis
#										- Results for biological process or cellular component analysis were exported as a
#										  tab-delimited file by clicking the blue "Table" button, which is the first export 
#										  option above the results (maternal_delivery_BPresults.txt, 
#										  maternal_delivery_CCresults.txt, fetal_delivery_BPresults.txt, 
#										  fetal_delivery_CCresults.txt)
#		-	will use Amigo database (http://amigo.geneontology.org/amigo):
#												- amigo_db.txt (downloaded on 05/12/2021)
#			note: will be used to match GO terms with maternal and fetal mtDEG-INTXs expression. Amigo file has entire
#				  entire Amgio database, comprised of GO accession IDs and terms for biological process GO terms, cellular
#				  component GO terms, and molecular function GO terms. Amigo file was generated as follows:
#										- Create blank text file (amigo_db.txt)
#										- Navigate to drill-down browser (http://amigo.geneontology.org/amigo/dd_browse)
#										- Click on grey bubble to right of GO term category ("biological_process",
#										  "cellular_component", "molecular_function") with total terms in white text.
#										- Click "retrieve gene products annotated to this term for this filter set"
#										  (hyperlinked text) at bottom of dialog box that pops up
#										- Specify "Homo sapiens" under "Organism" 
#										- Click on "Custom DL (up to 1000000)" button located above terms listed
#										- Move default options under "Selected fields" to "Available pool". Move
#										  "Acc (id)" and "Direct annotation (annotation_class_list)" from "Available pool"
#										  to "Selected fields"
#										- Click "Download" at bottom of dialog box
#										- Copy and paste all terms to amigo_db.txt text file
#
# OVERVIEW OF PROCESS:
#		1) COLLECT ALL UNIPROT IDs ASSOCIATED WITH ENRICHED GO TERMS
#		2) MATCH ENRICHED GO TERMS TO mtDEG-INTXs IDENTIFIED
#		3) MAKE PLOTTING FILES (MATERNAL AND FETAL AT DELIVERY)



###################################################
### 		###### READ IN FILES ######			###
###################################################

## READ IN AMIGO DATABASE ##
# [0] = Uniprot ID (in format UniProtKB:XXX)
# [1] = Gene name
# [2] = GO accession IDs associated with Uniprot ID, pipe ("|") delimited
amigo = open("amigo_db.txt", "r").read().split("\n")

## READ IN mtDEG-INTXs EXPRESSION FILES ##
# [0] = Ensembl transcript ID
# [1] = baseMean
# [2] = log2FoldChange
# [3] = lfcSE
# [4] = stat
# [5] = pvalue
# [6] = padj
# [7] = Uniprot_ID
maternal_expression = open("maternal_del_mtDEG-INTXs_expression.csv", "r").read().split("\n")

# [0] = Gene name
# [1] = DESeq2 (BH adjusted p-value)
# [2] = edgeR (BH adjusted p-value)
# [3] = ALDEx2 (effect size)
# [4] = Direction of expression
# [5] = Uniprot ID
fetal_expression = open("fetal_del_mtDEG-INTXs_expression.csv", "r").read().split("\n")

## READ IN GO ENRICHMENT FILES ##
# note: remove blank rows and informational rows
# [0] = GO term (term (GO:XXX))
# [1] = size of GO term (total genes belonging to GO term)
# [2] = total mtDEG-INTXs provided that belong to GO term
# [3] = proportion of provided mtDEG-INTXs that make up size of GO term
# [4] = fold enrichment
# [5] = raw P-value
# [6] = FDR (adjusted P-value)
maternal_BP = open("maternal_delivery_BPresults.txt", "r").read().split("\n")[11:]
maternal_CC = open("maternal_delivery_CCresults.txt", "r").read().split("\n")[11:]
fetal_BP = open("fetal_delivery_BPresults.txt", "r").read().split("\n")[11:]
fetal_CC = open("fetal_delivery_CCresults.txt", "r").read().split("\n")[11:]


################################################################
### 		###### GO ENRICHMENTS DICTIONARIES ######		 ###
################################################################

# make maternal GO enrichments dictionary
maternal_GOterms = {"BP":{}, "CC":{}}
for line in maternal_BP[1:]:
	if len(line) > 1:
		maternal_GOterms["BP"][line.split("\t")[0].split("(")[1][:-1]] = {"count":line.split("\t")[2],
		"pval":line.split("\t")[6],
		"GO_term":line.split("\t")[0].split("(")[0][:-1],
		"Uniprot_IDs":[]}
for line in maternal_CC[1:]:
	if len(line) > 1:
		maternal_GOterms["CC"][line.split("\t")[0].split("(")[1][:-1]] = {"count":line.split("\t")[2],
		"pval":line.split("\t")[6],
		"GO_term":line.split("\t")[0].split("(")[0][:-1],
		"Uniprot_IDs":[]}

# make fetal GO enrichments dictionary
fetal_GOterms = {"BP":{}, "CC":{}}
for line in fetal_BP[1:]:
	if len(line) > 1:
		fetal_GOterms["BP"][line.split("\t")[0].split("(")[1][:-1]] = {"count":line.split("\t")[2],
		"pval":line.split("\t")[6],
		"GO_term":line.split("\t")[0].split("(")[0][:-1],
		"Uniprot_IDs":[]}
for line in fetal_CC[1:]:
	if len(line) > 1:
		fetal_GOterms["CC"][line.split("\t")[0].split("(")[1][:-1]] = {"count":line.split("\t")[2],
		"pval":line.split("\t")[6],
		"GO_term":line.split("\t")[0].split("(")[0][:-1],
		"Uniprot_IDs":[]}

# populate GO enrichments dictionary with all possible Uniprot IDs associated with GO accession
for key, value in maternal_GOterms["BP"].items():
	for line in amigo:
		if key in line and "UniProtKB" in line.split("\t")[0]:
			if line.split("\t")[0].split(":")[1] not in value["Uniprot_IDs"]:
				value["Uniprot_IDs"].append(line.split("\t")[0].split(":")[1])				
print("Maternal BP Uniprot IDs collected")
for key, value in maternal_GOterms["CC"].items():
	for line in amigo:
		if key in line and "UniProtKB" in line.split("\t")[0]:
			if line.split("\t")[0].split(":")[1] not in value["Uniprot_IDs"]:
				value["Uniprot_IDs"].append(line.split("\t")[0].split(":")[1])	
print("Maternal CC Uniprot IDs collected")

for key, value in fetal_GOterms["BP"].items():
	for line in amigo:
		if key in line and "UniProtKB" in line.split("\t")[0]:
			if line.split("\t")[0].split(":")[1] not in value["Uniprot_IDs"]:
				value["Uniprot_IDs"].append(line.split("\t")[0].split(":")[1])				
print("Fetal BP Uniprot IDs collected")
for key, value in fetal_GOterms["CC"].items():
	for line in amigo:
		if key in line and "UniProtKB" in line.split("\t")[0]:
			if line.split("\t")[0].split(":")[1] not in value["Uniprot_IDs"]:
				value["Uniprot_IDs"].append(line.split("\t")[0].split(":")[1])	
print("Fetal CC Uniprot IDs collected")


########################################################
### 		###### FINAL DICTIONARIES ######		 ###
########################################################

# assign GO enrichment results to maternal mtDEG-INTXs identified
maternal_plotting_data = {"BP":{}, "CC":{}} 
for line in maternal_expression[1:]:
	for key, value in maternal_GOterms["BP"].items():
		if len(line) > 1 and line.split(",")[7] in value["Uniprot_IDs"]:
			maternal_plotting_data["BP"][line.split(",")[0] + "," + key] = {"Expression":line.split(",")[2],
			"pval":value["pval"], "count":value["count"], "GO_term":value["GO_term"], "Uniprot_ID":line.split(",")[7]}
	for key, value in maternal_GOterms["CC"].items():
		if len(line) > 1 and line.split(",")[7] in value["Uniprot_IDs"]:
			maternal_plotting_data["CC"][line.split(",")[0] + "," + key] = {"Expression":line.split(",")[2],
			"pval":value["pval"], "count":value["count"], "GO_term":value["GO_term"], "Uniprot_ID":line.split(",")[7]}

# assign GO enrichment results to maternal mtDEG-INTXs identified
fetal_plotting_data = {"BP":{}, "CC":{}} 
for line in fetal_expression[1:]:
	for key, value in fetal_GOterms["BP"].items():
		if len(line) > 1 and line.split(",")[5] in value["Uniprot_IDs"]:
			fetal_plotting_data["BP"][line.split(",")[0] + "," + key] = {"Expression":line.split(",")[4],
			"pval":value["pval"], "count":value["count"], "GO_term":value["GO_term"], "Uniprot_ID":line.split(",")[5]}
	for key, value in fetal_GOterms["CC"].items():
		if len(line) > 1 and line.split(",")[5] in value["Uniprot_IDs"]:
			fetal_plotting_data["CC"][line.split(",")[0] + "," + key] = {"Expression":line.split(",")[4],
			"pval":value["pval"], "count":value["count"], "GO_term":value["GO_term"], "Uniprot_ID":line.split(",")[5]}


########################################################
### 		###### WRITE PLOTTING FILES ######		 ###
########################################################

# note: z-scores are calculated in R. Column names meant to be compatible with GOplot package (R)

# maternal file
maternal_plotting_file = open("maternal_delivery_GOplot_data.csv", "w")
maternal_plotting_file.write("category,Ensembl ID,ID,term,count,genes,logFC,adj_pval\n")

for key, value in maternal_plotting_data["BP"].items():
	maternal_plotting_file.write("BP" + "," + key + "," + value["GO_term"] + 
	"," + value["count"] + "," + value["Uniprot_ID"] + "," + value["Expression"] + "," + value["pval"] + "\n")
for key, value in maternal_plotting_data["CC"].items():
	maternal_plotting_file.write("CC" + "," + key + "," + value["GO_term"] + 
	"," + value["count"] + "," + value["Uniprot_ID"] + "," + value["Expression"] + "," + value["pval"] + "\n")

# fetal file
fetal_plotting_file = open("fetal_delivery_GOplot_data.csv", "w")
fetal_plotting_file.write("category,Gene name,ID,term,count,genes,logFC,adj_pval\n")

for key, value in fetal_plotting_data["BP"].items():
	fetal_plotting_file.write("BP" + "," + key + "," + value["GO_term"] + 
	"," + value["count"] + "," + value["Uniprot_ID"] + "," + value["Expression"] + "," + value["pval"] + "\n")
for key, value in fetal_plotting_data["CC"].items():
	fetal_plotting_file.write("CC" + "," + key + "," + value["GO_term"] + 
	"," + value["count"] + "," + value["Uniprot_ID"] + "," + value["Expression"] + "," + value["pval"] + "\n")

print("Files written. Script complete")