Datasets
Models
140 lines (107 with data), 5.5 kB
Differential Expression Analysis
InMoose offers a Python port of the well-known R Bioconductor packages:
- DESeq2 package [Love2014]_ in module :doc:`deseq`.
- edgeR package [Chen2016]_ in module :doc:`edgepy`.
- limma package [Ritchie2015]_. in module :doc:`limma`.
Note that not all features of the R packages are necessarily ported. Extending the functionality of these modules will be based on user requests, so do not hesitate to open an issue if your favorite feature is missing.
In addition, InMoose provides a meta-analysis feature to combine the results from different differential expression analysis tools.
Please refer to [Colange2024] for a detailed comparison of InMoose implementation with the original R implementations.
Differential Expression Meta-Analysis
We illustrate the differential expression meta-analysis capabilities of InMoose along two approaches:
- the Aggregate Data (AD) approach consists in running classical differential expression tools on individual cohorts then combining the results through e.g. random-effect models.
- the Individual Sample Data (ISD) consists in merging individual cohorts into a large meta-cohort, accounting for batch effects to eliminate inter-cohort biases, then running a classical differential expression analysis on the resulting meta-cohort.
We start by simulating RNA-Seq data, using the :mod:`sim` module of InMoose.
We then run the two meta-analysis approaches on the obtained data.
We can now compare the results obtained by the two approaches.
It is possible to combine results obtained from different tools, as long as the results of the differential expression analysis are stored as :class:`DEResults`. All three modules :doc:`limma`, :doc:`edgepy` and :doc:`deseq` return sub-classes of :class:`DEResults`, thus allowing users to perform cross-technology meta-analysis (e.g. by combining results from :doc:`limma` with results from :doc:`deseq`).
References
| [Colange2024] | M. Colange, G. Appé, L. Meunier, S. Weill, A. Nordor, A. Behdenna. 2024. Differential Expression Analysis with InMoose, the Integrated Multi-Omic Open-Source Environment in Python. Bioarxiv. :doi:`10.1101/2024.11.14.623578` |