nfkb-tag-seq

The following readme is the code used for processing and analysis of TAGseq data in "The Effects of IKKβ Inhibition on Early NF-κB Activation and Transcription of Downstream Genes". The link to the article is provided here: https://www.sciencedirect.com/science/article/pii/S0898656818303061. In addition, the abstract is provided below.

Abstract

Small molecule approaches targeting the nuclear factor kappa B (NF-kB) pathway, a regulator of inflammation, have thus far proven unsuccessful in the clinic in part due to the complex pleiotropic nature of this network. Downstream effects depend on multiple factors including stimulus-specific temporal patterns of NF-kB activity. Despite considerable advances, genome-level impact of changes in temporal NF-kB activity caused by inhibitors and their stimulus dependency remains unexplored. This study evaluates the effects of pathway inhibitors on early NF-κB activity and downstream gene transcription. 3T3 fibroblasts were treated with SC-514, an inhibitor targeted to the NF-kB pathway, prior to stimulation with interleukin 1 beta (IL-1β) or tumor necrosis factor alpha (TNF-α). Stimulus induced NF-κB activation was quantified using immunofluorescence imaging over 90-minutes and gene expression tracked over 6-hours using mRNA TagSeq. When stimulated with IL-1β or TNF-α, significant differences (P < 0.05, two-way ANOVA), were observed in the temporal profiles of NF-κB activation between treated and untreated cells. Increasing numbers of differentially expressed genes (P < 0.01) were observed at higher inhibitor concentrations. Individual gene expression profiles varied in an inhibitor concentration and stimulus-dependent manner. The results in this study demonstrate small molecule inhibitors acting on pleiotropic pathway components can alter signal dynamics in a stimulus-dependent manner and affect gene response in complex ways.

This code relies on the code used for the original TagSeq analysis paper located at https://github.com/z0on/tag-based_RNAseq. The original paper can be found here:

Dependencies

Processing

Command line interface. Perl.

Analysis

Software	Version
R version 3.5.1 (2018-07-02)
tidyverse
dplyr
tibble
gplots
RColorBrewer
genefilter
calibrate

library(RColorBrewer) library(gplots) library(genefilter) library(calibrate) library(DESeq2) library(data.table) library(wesanderson)

Sequencing data format

This code is also generated using the file format for reads downloaded from Illumina's BaseSpace Sequence Hub: https://login.illumina.com/platform-services-manager/?rURL=https://basespace.illumina.com&clientId=basespace&clientVars=aHR0cHM6Ly9iYXNlc3BhY2UuaWxsdW1pbmEuY29tL2Rhc2hib2FyZA&redirectMethod=GET#/. The original data can be downloaded here: https://www.ncbi.nlm.nih.gov/sra?term=SRP163157. This repository will be updated with scripts to analyze the data in SRA format after installing sra-toolkit (https://ncbi.github.io/sra-tools/install_config.html) as follows:

$ fastq-dump SRP163157

For now, data in BaseSpace format can be requested from the original authors if needed (meghanbloom@utexas.edu).

Processing TagSeq data

First, clone this repository:

$ git clone https://github.com/sachitsaksena/nfkb-tag-seq.git
$ cd nfkb-tag-seq

Move the BaseSpace downloaded TagSeq data to this repository:

$ mv /path/to/Tag/Seq/data .

Then run the following script while still in the nfkb-tag-seq directory to organize reads by experimental condition and concatenate lanes of each RNAseq run:

$ bash concatenate_lanes.sh

Then clone the original repository while stil in the same directory for access to processing and clipping scripts scripts:

$ git clone https://github.com/z0on/tag-based_RNAseq

Now, to process simply run the following set of commands within the same directory:

First generate a set of commands that trims reads and run the process in "screen," which will take some time:

$ ./tagseq_trim_launch.pl '\.fastq$' > clean
# open a screen session
$ screen
# now in screen run
$ . clean
# ctrl + A + D to exit screen while still runnning process
# exit screen
$ exit

Then generate a set of commands to map to a reference genome, which will take much more time

$ ./tagseq_bowtie2map.pl "trim$" Mus_musculus.GRCm38.cdna.all.fa > maps
# open screen
$ screen
# now in screen
$ . maps
# ctrl + A + D to exit screen while still runnning process
# exit screen
$ exit

Now, run the following command:

$ grep '^>' Mus_musculus.GRCm38.cdna.all.fa |\
 perl -pe 's/>(\S+)\s.*gene:(\S+).*/\1\t\2/g' >\
 Mus_musculus.GRCm38.cdna.all_seq2iso.tab

Next, run samcounts via a perl script in the original author's repository:

$ ./samcount_launch_bt2.pl '\.sam' Mus_musculus.GRCm38.cdna.all_seq2iso.tab > sc
$ screen
$ . sc
## ctrl + A + D to exit screen while still runnning process
## "exit" to exit screen and kill process
$ exit

Finally, compile expression measurements from all the disparate samples:

$ ./expression_compiler.pl *.sam.counts > allcounts.txt

A special thanks to Dennis Wylie at UT Austin for help with the command generating scripts.

Analysis reproduction

R scripts can be run from start to finish on the data produced from processing above. check_replicates.R is quality control code for ensuring high quality data by evaluating clustering of biological replicates.

diff_exp_analysis.R is differential expression analysis used in the paper.