This repository serves as the main resource to reproduce our results from the benchmarking study of second-generation deconvolution methods.
The repository is divided into three sections:
1) docker/
Here we store the Dockerfile that can be used in the benchmarking pipeline. It includes the [omnideconv](https://github.com/omnideconv/omnideconv) R package with all of its dependencies. An image created by this Dockerfile can also easily be accessed from [Dockerhub](https://hub.docker.com/repository/docker/alexd13/omnideconv_benchmark/general). 2) pipeline/
This folder includes the nextflow pipeline that can be used to reproduce all results from our study and can also be extended to benchmark new deconvolution methods. A detailed description is included in the directory itself.3) visualization
This folder includes the scripts that generate the main figures of our study. The directly use the results of the benchmarking pipeline as input.