Contains scrips and files used for the preprocessing steps.
Scripts to download the raw fastq files from a given list of GEO IDs.
Scripts and output to counts the number of biological replicates in the dataset.
Scripts and information generated to look for duplicated sequences in the dataset.
Scripts used to generated the sample-matrix used as an input for the nf-core/chip-seq pipeline.
File containing information about all the samples used in the analysis. Contains files names as well as information about time stamp (e.g. L1, P13, I10), experiment type (ChIP-seq, RNA-Seq), antibody (for ChIP-seq only) and comments on the file (use as control, duplicated, etc.).