#!/usr/bin/env python
# coding: utf-8

# In[1]:


import os
import pandas


# In[ ]:





# In[2]:


path = os.path.join('data', 'problems.tsv')
problem_df = pandas.read_table(path)
problem_df['name_lower'] = list(map(str.lower, problem_df.problem))
problem_df[:3]


# In[3]:


len(problem_df)


# In[4]:


url = 'http://git.dhimmel.com/disease-ontology/data/term-names.tsv'
doterm_df = pandas.read_table(url)
doterm_df['name_lower'] = list(map(str.lower, doterm_df.name))
set(doterm_df.type)


# In[5]:


doterm_df = doterm_df.query("type in ['exact-synonym', 'name']")
doterm_df[:3]


# In[6]:


domap_df = problem_df.merge(doterm_df, how='left', on='name_lower')
path = os.path.join('data', 'problem-to-doid.tsv')
domap_df.to_csv(path, index=False, sep='\t')


# In[7]:


pair_df = domap_df[['problem_definition_id', 'doid']].dropna().drop_duplicates()
len(pair_df)


# In[8]:


domap_df[:50]


# In[ ]:





# In[9]:


mapped_problems = set(pair_df.problem_definition_id)


# In[10]:


len(mapped_problems) / len(problem_df)


# In[18]:


len(mapped_problems)


# In[11]:


# Find problems that mapped to multiple DO terms
duplicates = set(pair_df[pair_df.duplicated('problem_definition_id')].problem_definition_id)
domap_df[domap_df.problem_definition_id.isin(duplicates)]


# In[ ]:





# In[16]:


url = 'http://git.dhimmel.com/disease-ontology/data/slim-terms.tsv'
doslim_df = pandas.read_table(url)
len(doslim_df)


# In[17]:


sum(doslim_df.doid.isin(set(pair_df.doid)))


# In[24]:


# with propagation
url = 'http://git.dhimmel.com/disease-ontology/data/slim-terms-prop.tsv'
doslim_df = pandas.read_table(url)
len(set(doslim_df.slim_id[doslim_df.subsumed_id.isin(set(pair_df.doid))]))


# In[ ]: