VARS-seq

Designing sequence libraries for VARS-seq and analyzing gDNA and RNA sequencing data from VARS-seq.

Overview

This repository contains scripts for:

• Designing structure variant sequences for VARS-seq in library_design/
  • Sampling random variants and rescue sequences for candidate stem disruption and rescue sequences
  • Predicting secondary structures for sampled sequences
  • Scoring secondary structures for desired properties for disrupting and rescuing stem sets
  • Utilities for processing RNA secondary structures
• Analyzing gDNA sequencing data for VARS-seq in gdna_analysis/
  • Pairing barcodes with consensus sequences
  • Obtaining transformation frequencies per barcode
  • Obtaining barcodes for designed variant sequences
  • Plotting designed variant sequence coverage and length distributions for consensus sequences
  • Filtering out barcodes with multiple consensus sequences
• Analyzing RNA sequencing data for VARS-seq in rnaseq_analysis/
  • Obtaining per-barcode, per-variant, and per-stem set counts for spliced and unspliced reads after deduplicating UMI's
  • Displaying heatmaps showing the significance of differences between wildtype, variant, and rescue sequence sets
  • Assembling stats tables for the number of spliced and unspliced reads per barcode, grouped by variant sequence

Designing sequences for VARS-seq

Sample commands for designing variant library sequences:

• First sample candidate variant sequences and predict their secondary structures:
  • python library_design/generate_variant_mfes.py library_design/intron_info/rps9b/rps9b.fasta library_design/intron_info/rps9b/rps9b_secstruct.txt library_design/intron_info/rps9b/rps9b_bpp.csv library_design/output_mfes/rps9b_varmfe_ 238 10000 library_design/intron_info/rps9b/rps9b_var_list.txt
• Then score candidate variant sequences:
  • python library_design/score_variants.py library_design/intron_info/rps9b/rps9b.fasta library_design/intron_info/rps9b/rps9b_secstruct.txt library_design/intron_info/rps9b/rps9b_bpp.csv library_design/intron_info/rps9b/rps9b_var_list.txt library_design/output_mfes/rps9b_varmfe_ library_design/barcodes/barcodes_n8_k3.20220511_104258.csv library_design/intron_info/rps9b/rps9b_fullseq.txt library_design/variants/rps9b_variants.csv 4

Detailed descriptions for the input and output variables and data types for these commands are included in the library_design/ directory's README.

We recommend carrying out sequence design with the help of a high performance cluster. To accelerate sequence design, we carried out secondary structure prediction and scored secondary structures for each stem set in each intron on separate cores in parallel.

Producing figures

Each subdirectory includes a figures directory with example analysis figures and a README detailing how to produce these figures.

Requirements and installation

• scipy v1.10.1
• numpy v1.24.3
• matplotlib v3.7.1
• seaborn v0.12.2
• pandas v2.0.2
• arnie: https://github.com/DasLab/arnie
• Contrafold 2.00: http://contra.stanford.edu/contrafold/download.html

Installation: Install all requirements (expected time: 1-2 hours) and then clone this repo.