def get_pathway_protein_names(self, pathway):
"""returns list of list. Each elements is made of 3 items: gene name,
locusId and accession (often empty
Requires to parse HTML page such as
http://www.biocarta.com/pathfiles/m_actinYPathway.asp
to figure out the URL that would pop up if we press the protein list
button. For instance:
http://www.biocarta.com/pathfiles/PathwayProteinList.asp?showPFID=175
but now we need to parse the HTML, which is not necessaray very robust.
THere are many tables and we want to access one that is a children of
another... Finally, We scan the table for tr and td tags.
The most difficult is to find the good table which is hardcoded to be
the third that contains a th[0] == "Gene name". Although there is only
one, 3 are returned due probably to an error in the parsing or the HTMl
file itself. To be checked and made more robust.
"""
url = self._url + "/pathfiles/" + pathway
x = readXML(url)
self.logging.info("Reading " + url)
protein_url = [this.get("href") for this in x.findAll("a") \
if 'href' in this and "Protein" in this.get("href")]
if len(protein_url) == 0:
return None
else:
link = protein_url[0]
link = link.split("/pathfiles/")[-1]
link = str(link) # get rid of unicode ?
link = link.strip("')")
url = self._url + "/pathfiles/" + link
self.logging.info("Reading " + url)
x = readXML(url)
# seems to work
table = [
this for this in x.findAll("table") if this.findAll("th")
and this.findAll("th")[0].getText() == "Gene Name"
][2]
# now get the genename, locus and accession
rows = [[y.getText() for y in xx.findAll("td")]
for xx in table.findAll("tr")]
rows = [xx for xx in rows if len(x)]
return rows
def get_pathway_protein_names(self, pathway):
"""returns list of list. Each elements is made of 3 items: gene name,
locusId and accession (often empty
Requires to parse HTML page such as
http://www.biocarta.com/pathfiles/m_actinYPathway.asp
to figure out the URL that would pop up if we press the protein list
button. For instance:
http://www.biocarta.com/pathfiles/PathwayProteinList.asp?showPFID=175
but now we need to parse the HTML, which is not necessaray very robust.
THere are many tables and we want to access one that is a children of
another... Finally, We scan the table for tr and td tags.
The most difficult is to find the good table which is hardcoded to be
the third that contains a th[0] == "Gene name". Although there is only
one, 3 are returned due probably to an error in the parsing or the HTMl
file itself. To be checked and made more robust.
"""
url = self._url + "/pathfiles/" + pathway
x = readXML(url)
self.logging.info("Reading " + url)
protein_url = [this.get("href") for this in x.findAll("a") \
if 'href' in this and "Protein" in this.get("href")]
if len(protein_url) == 0:
return None
else:
link = protein_url[0]
link = link.split("/pathfiles/")[-1]
link = str(link) # get rid of unicode ?
link = link.strip("')")
url = self._url + "/pathfiles/" + link
self.logging.info("Reading " + url)
x = readXML(url)
# seems to work
table = [this for this in x.findAll("table") if this.findAll("th")
and this.findAll("th")[0].getText() == "Gene Name"][2]
# now get the genename, locus and accession
rows = [[y.getText() for y in xx.findAll("td")] for xx in table.findAll("tr")]
rows = [xx for xx in rows if len(x)]
return rows
def get_pathway_names(self, startswith=""):
"""returns pathways from biocarta
all human and mouse. can perform a selectiom
"""
x = readXML(self._allPathwaysURL)
pathways = [this.get("href") for this in x.findAll("a") if "pathfiles" in this.get("href")]
pathways = [str(xx.split("/")[-1]) for xx in pathways] # split the drive
pathways = sorted(list(set(pathways)))
pathways = [xx for xx in pathways if xx.startswith(startswith)]
return pathways
def get_pathway_names(self, startswith=""):
"""returns pathways from biocarta
all human and mouse. can perform a selectiom
"""
raise NotImplementedError
allx = readXML(self._allPathwaysURL)
pathways = [
this.get("href") for this in allx.findAll("a")
if "pathfiles" in this.get("href")
]
pathways = [str(x.split("/")[-1]) for x in pathways] # split the drive
pathways = sorted(list(set(pathways)))
pathways = [x for x in pathways if x.startswith(startswith)]
return pathways
