This code is associated with the paper from Zhou et al., "Codon usage biases co-evolve with transcription termination machinery to suppress premature cleavage and polyadenylation". eLife, 2018. http://dx.doi.org/10.7554/eLife.33569

poly(A)-seq pipeline

These codes were written to analyze PCPA (prematural cleavage and polyadenylation) at the 3' UTR, intron and CDS. For details of the analyses, please refer to our paper titled "Codon usage biases co-evolve with the transcription termination machinery to suppress premature cleavage and polyadenylation in coding regions".

The use of arguments were written in the each scripts and can be changed for your need. To replicate the results in our publication, please use the settings described in "Materials and Methods", if settings in scripts are different.

Before start

require Perl(5.22) and R(3.4.0) or above

1, download and install bowtie2, tophat2, SAMTOOLS, BEDTools.

make sure the installed softwares are in your PATH.

2, create the reference files (ORF sequences, CAI and CBI values).

Please go to iGenome (Illumnia) or other sources to download your desired genomic references, including genome fasta file, GTF and refFlat.

use the perl script to create ORF reference sequences.

open the script sequence_bu_refFlat.pl, make sure to give it the right path to access reference files. The run the code in termial

perl sequence_by_refFlat.pl 

Then put script CBI_ref_define.R and cal_CBIAvg.pl inside the ref folder. excute the R scripts in Rstudio and Perl in terminal as showed above to create CAI and CBI reference file.

Note: install necessary module for R and Perl used in the codes

For R: ggplot2 reshape2

For Perl: List::MoreUtil List::Compare Math::CDF

3, put all your fastq files in a folder named "fq". Then download all scripts from "3READ" folder and place them outside the "fq" folder

step1 to step4， processing reads and mapping.

change the file name according inside the scripts (Shell or Perl), with a text editor. Then execute these code as below:

perl step1_trim3.pl
sh step2_align.sh
perl step3_PASS.pl
sh step4_position.sh

The above codes were written for 3READS sequencing method. If 2P-seq are used, please use the codes followed by *_2P-seq.pl

step5

This step will create:

1. scatter plots of CBI/CAI to normalized ORF/3'UTR termination ratio
2. line plots showing the GC content flanking the pA sites within ORFs or 3' UTRs HOW TO USE: change the file name inside the shell script and run, which should produce figures and files in subfolders

sh step5_PAS_search.pl

step6

This step will create: codon and dicodon frequency around the putative PAS region.
HOW TO USE: open the shell script and change the file names accordingly. Then excute as following to analyze codon and codon frequencies

sh step6_PAS_dicodon.sh

then execute step6_PAS_dicodon.R in Rstudio software.

step7

This step will create: 1.The line plot of AU content flanking selected 6nt PAS motifs. 2. relative codon refrequency flanking the PAS motifs. HOW TO USE: Open and change the file name accordingly. Then simply execute the shell script, which will produce the figures

sh step7_PAS_prediction.sh

step8

This step will create the PSSM scores for PAS region of 3'UTR and ORF regions. execute the following script to calculate the PAS scores based on PSSM

perl step8_PSSM.pl

To make the boxplot, simly use the built-in function "boxplot" in R.

---

Please direct all your questions to Yunkun Dang (email: yunkun_dang@126.com)

Physical address: Center for Life Science, College of Biological Sciences, Yunnan University, Kunming, Yunnan, China.