Helpful scripts for performing routine analyses
This is an explanation of how to use the helper scrpts in this repo to go from a folder with the feature_counts output to a basic DE analysis
Basically bedtools intersect but in R. Script that can take an annotation file (GTF/BED, or can use inbuilt annotatr genic annotations) and BED file containing coordinates of features (e.g. iCLIP sites) and report where feature coordinates intersect with annotation coordinates. Outputs tab-separated tablewhere each line contains feature info and info of annotation which which it overlaps. Can also filter for overlaps by 'feature field' (3rd column/field) in provided GTF file (e.g. gene,exon,intron,UTR etc.).
Run Rscript feature-annotation-overlap.R -h at the command line to get more info on options.
(Stay tuned might add more functionality)
Follows on from check_rg_header.sh. If you ran header script in directory containing soft links to the faulty BAMs (very sensible), then this script can match up the fixed header file with original location of the unfixed BAM the corresponding soft link points to. The script then moves the fixed bam to overwrite the faulty BAM in its original location.
bash mv_fixed_bams_to_soft_link.sh <-h/--help> for usage info
Bash script that looks through directory and subdirectories for BAM files that are missing @RG flag at start of read group header and adds @RG accordingly. Used on some older processed data thathad these missing so re-processed BAMs can be used with more recent versions of samtools, pysam etc.
Takes a basic BED file and converts it into a GTF format which can be used by featureCounts
Takes a folder of aligned bams (right now pretty inflexibly) and runs the picard tools collect insertsize metrics tools on them
Take a data table or data frame and returns a list of a data frame with rownames as genes, and a metadata table for use with the DESEQ2 R tool
Takes a folder path with featureCount tables per sample and makes it into a tidy table
Helper R functions for converting between mouse/human(Using bioMart) and converting between ENTREZ,HUGO,ENSEMBL ids using org.Annotation.dbs
Functions which read the STARLogFiles and FeatureCountsMapped files to retrieve mapping statistics
Imports the functions from librarySize and creates a csv with mapping stats based on the variables in the script. Currently a kludge solution
Streamline feature-annotation-overlap.R command line options Test feature-annotation-overlap.R with -d option to use inbuilt annotations Fix collect_input_stat to operate more flexibly Mapping_stat.py should read from the command line not global variables