polish code, add README and requirements.txt

Makefile

@@ -0,0 +1,28 @@
+.DEFAULT_GOAL := analysis
+
+requirements:
+	pip install -r requirements.txt
+
+makeref:
+	python src/guides_to_ref.py
+
+# process project's data
+# with looper/pypiper/pipelines:
+# see https://github.com/epigen/looper
+# see https://github.com/epigen/pypiper
+# see https://github.com/epigen/pipelines
+process:
+	looper run metadata/config.yaml
+
+assign:
+	python assign_gRNA_cells.py
+
+collect:
+	python src/collect_expression.py
+
+analysis: assign collect
+	python src/analysis.py
+
+all: requirements makeref process assign collect analysis
+
+.PHONY: requirements makeref process assign collect analysis all

README.md

@@ -1 +1,32 @@
-# crop-seq
+## Pooled CRISPR screening with single-cell transcriptome read-out
+
+Paul Datlinger, André F Rendeiro<sup>\*</sup>, Christian Schmidl<sup>\*</sup>, Thomas Krausgruber, Peter Traxler, Johanna Klughammer, Linda C Schuster, Amelie Kuchler, Donat Alpar, Christoph Bock. Nature Methods (2017).
+
+<sup>\*</sup>These authors contributed equally to this work
+
+**Paper**: [http://dx.doi.org/](http://dx.doi.org/)
+
+**Website**: [crop-seq.computational-epigenetics.org](http://crop-seq.computational-epigenetics.org)
+
+This repository contains scripts used in the analysis of the data in the paper.
+
+<br>
+
+### Analysis
+
+In the [paper website](http://crop-seq.computational-epigenetics.org) you can find most of the output of the whole analysis.
+
+Here are a few steps needed to reproduce it:
+
+1. Clone the repository: `git clone [email protected]:epigen/crop-seq.git`
+2. Install required software for the analysis:`make requirements` or `pip install -r requirements.txt`
+
+If you wish to reproduce the processing of the raw data (all data has been deposited at [GEO](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE92872)), run these steps:
+
+1. Download the data localy.
+2. Prepare [Looper](https://github.com/epigen/looper) configuration files similar to [these](metadata/config.yaml) that fit your local system.
+3. Prepare a genome annotation containing gRNA sequences `make makeref` and adapt the pipeline [configuration file](metadata/pipeline_config.yaml) to point to the created files.
+4. Run samples through the pipeline: `make preprocessing` or `looper run metadata/config.yaml`
+5. Run the analysis: `make analysis`
+
+Additionaly, gene expression matrices containing cell metadata (including gRNA assignments) are available from [GEO with accession number GSE92872](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE92872).