Disease cell atlas approximations - API

This repository provides an API to fetch and analyze cell atlas data on various human diseases. For the time being, the API uses data from Cellxgene Census. Single cell omic data is compressed into h5 files using scquill.

---

NOTE

Compared to Cellxgene itself, this API is designed to be much faster and closer to typical biomedical questions.

---

Features

Currently implemented:

• Analyze differential cell type abundance across disease states, sexes, and developmental stage.
• Analyze differential gene expression for all cell types or specific cell types in disease contexts.
• Fetch metadata related to diseases, including datasets, organs, and conditions.
• Retrieve average gene expression for specific genes across disease states.
• Identify top differentially expressed genes in disease-related contexts (e.g., kidney diseases).

Documentation

Tutorial and reference documentation is available at https://cell-atlas-approximations-disease-api.readthedocs.io/en/latest/.

Usage

REST

REST

The REST interface is language-agnostic and can be queried using any HTTP request handler, e.g. in JavaScript:

(async () => {
  let response = await fetch("https://api-disease.atlasapprox.org/v1/metadata?disease=covid&cell_type=alveolar type 2");
  if (response.ok) {
    let data = await response.json();
    console.log(data);
  }  
})();

Python

Python

The Python interface uses a central API class. Its methods implement the REST endpoints:

import atlasapprox_disease

api = atlasapprox_disease.API()
print(api.metadata())
print(api.average(disease="covid", features="IGHG1,CXCL13,S100A8"))

JavaScript

JavaScript/nodejs

An object containing one function for each API endpoint is exported by the atlasapprox npm package:

let atlasapprox = require('atlasapprox-disease');
(async () => {
  let data = await api.metadata(disease:"covid");
  console.log(data);
  }  
)();