This repository contains scripts for meta-analysis of four studies which observed that the gut microbiome composition plays a large role in predicting the treatment outcome of patients undergoing cancer immunotherapy.
- Last update: 2021-09-21
This repository is administered by Angelo Limeta (@angelolimeta), Division of Systems and Synthetic Biology, Department of Biology and Biological Engineering, Chalmers University of Technology.
This project concerns five published papers:
- Gopalakrishnan et al. 2018, Science
- Matson et al. 2018, Science
- Routy et al. 2018, Science
- Frankel et al. 2018, Neoplasia
- Peters et al. 2019, Genome Medicine
All studies utilize metagenomic sequencing through either Whole Genome Shotgun (WGS) or 16S rRNA.
| Study | Responders | Non-responders | Total | Sequencing Method | Database | ID | Usage in meta analysis |
|---|---|---|---|---|---|---|---|
| Gopalakrishnan et al. | 14 | 11 | 25 | Illumina HiSeq 2000 | ENA | PRJEB22893 | Discovery |
| Matson et al. | 16 | 26 | 42(39) | Illumina NextSeq 500 | SRA | PRJNA399742 | Discovery |
| (Routy et al.) | ? | ? | 127 | Thermofisher Ion Proton | ENA | PRJEB22863 | Not used |
| Frankel et al. | 24 | 9 | 33 | Illumina HiSeq 2000 | SRA | PRJNA397906 | Discovery |
| Peters et al. | 18 | 9 | 27 | Illumina HiSeq 2500 | SRA | PRJNA397906 | Validation |
In the Peters et al. data set, response is defined as PFS >= 6 mo. for R and PFS < 6 mo. for NR. Only WGS samples taken prior to immune checkpoint inhibition are used. For example, the Routy et al. data set included several WGS samples from the same patient at different timepoints during treatment (T0,T1,T2). However, only the T0 samples are used.
Treatment response is evaluated through the Response Evaluation Criteria in Solid Tumors 1.1 (RECIST 1.1). Patients with complete or partial response, according to RECIST 1.1 criteria, are classified as responders; whereas patients with stable or progressive disease are classified as non-responders.