RDP Web Client

rdp_web_client.py

import requests
import pandas as pd
from tqdm import tqdm
from os.path import join
from functools import wraps

def unpaginate(endpoint):
    """Unpaginate a paginated Spring endpoint by yielding from the pages content"""
    @wraps(endpoint)
    def f(*args, page_size=20):
        page = endpoint(*args, page_size=page_size)
        total_pages = page['totalPages']
        yield from page['content']
        for p in range(1, total_pages):
            yield from endpoint(*args, page=p, page_size=page_size)['content']
    return f

class Api:
    def __init__(self, url, token=None, session=None):
        self.url = url
        self.session = requests.Session() if session is None else session
        if token is not None:
            self.session.headers.update({'Authorization': f'Bearer {token}'})
        
    def _request(self, endpoint, **kwargs):
        res = self.session.get(join(self.url, endpoint), params=kwargs)
        res.raise_for_status()
        return res.json()
    
    @unpaginate
    def get_users(self, page=0, page_size=20):
        return self._request('api/users', page=page, size=page_size)
    
    @unpaginate
    def get_genes(self, page=0, page_size=20):
        return self._request('api/genes', page=page, size=page_size)
      
api = Api('https://register.rare-diseases-catalyst-network.ca/')

# retrieve all genes by consuming the generator
genes = list(tqdm(api.get_genes(page_size=500)))

genes_df = pd.DataFrame(genes)

# get the number of unique manually added genes
genes_df[genes_df.tier.isin(['TIER1', 'TIER2'])].geneId.unique().size

# get number of unique genes per taxon
genes_df['taxon_id'] = genes_df['taxon'].map(lambda t: t['id'])
genes_df.groupby('taxon_id').geneId.unique().map(len)