search

aryeelab / crispr_mutation_reanalysis

Commands, scripts, and data for reanalysis of "Unexpected mutations after CRISPR–Cas9 editing in vivo"
https://aryeelab.org/crispr_mutation_reanalysis
2 stars 1 forks source link
bioinformatics crispr-cas9 in-vivo mutations off-target

readme



“Unexpected mutations after CRISPR-Cas9 editing in vivo” are most likely pre-existing sequence variants and not nuclease-induced mutations

This web resource contains the scripts, commands, and most data required for what we reported in this bioRxiv preprint.

Our re-analysis was conducted by members of the Aryee, Joung, and Pinello Labs at MGH/Harvard in response to this manuscript.

Overview

In brief, we genotype the mice from the WGS data using GATK and replicate the "cancer pipeline" calls from the original informatic analysis pipeline. A succinct summary how we structured the sensible assumptions needed to determine if mutations were caused by Cas9 contrasted with the observed data are contained in our main figure.

Figure 1. Measures of genetic relatedness in the F03, F05 and FVB mice. Schaefer et al.’s experimental model to examine mutations attributable to Cas9 treatment relies on the assumption that mice are isogenic, meaning no private mutations are observed within or shared between mice as depicted in (A). However, after genotyping these mice using GATK best practices, we observe a significant departure from this model, suggesting that the F03 and F05 mice are more genetically related at common variants (B; n = 31,079). As an isogenic system is practically impossible, the selection of littermate assumes the number of loci with shared genotypes is nearly identical for all mice, leading to equivalent genetic distances separating them as shown in (C). This representation of an equal genetic model demonstrates a clear departure from the observed data at common variants and other non-dbSNP loci that we term “novel variants” (D; n = 38,981). Additionally, the variants previously reported by Schaefer et al. (dark gray, D) represent a small subset of the genotypes common to F03 and F05 but distinct from FVB at non-dbSNP sites. The observed ratios in B and D cannot be distinguished from each other (p = 0.304; two-sided Fisher’s Exact Test), but each represent a significant departure (p < 2.2 10-16; Chi-Squared Test) from the equal genetic distance model (C) required to attribute differential SNVs to Cas9 activity. Panel (E) provides a graphical summary of these models of genetic relatedness using under the two hypothetical models and the two sets of observed variants.

Data access

Arguably, the most important files for understanding our response manuscript are the raw genotype files called from GATK. These can be accessed from an AWS instance as the files exceed a combined 7 gigabytes. 

wget https://s3.amazonaws.com/crispr-mutation-reanalysis/allSamples.merged.fullData.vcf.gz
wget https://s3.amazonaws.com/crispr-mutation-reanalysis/allSamples.merged.fullData.vcf.gz.tbi

Additionally, the full set of loci that we used to annotate whether variants were in dbSNP or not:

wget https://s3.amazonaws.com/crispr-mutation-reanalysis/allSNPannotations.bed.gz

Note: this isn't a .bed file per se as it only has two columns, which was constructed as such to save space. A third column can be added using a simple awk command and adding 1 to the second field (i.e.: $2 +1) and potentially a dummy fourth column depending on the down-stream analysis tool.

Finally, the original authors were gracious enough to provide us with their code and results, which we have integrated into our analysis framework and hosted in our git repository here.

Our analyses

Below is a synopsis of what we did and how one can navigate this web resource to access the files and data.

Variant Recalling

The pipeline for recalling variants from the .fastq files is found here. Our reanalysis followed this protocol for best practices in aligning and prepping whole-genome sequencing samples.

Main text figure

The code for reproducing the main text figure can be found here. In brief, the absolute numbers of variants are divided into proportions to produce the four panels. The .bed files for getting this numbers (use wc -l *.bed) are here.

Supplement

Code for the workflow, figures, etc. associated with our supplement can be found here.

File naming conventions

Throughout the repository, most samples in the reanalysis are ID'd by the last two digits of the SRR number. This table is the most useful key--

FVB - Control - 98
F03 - Treated - 97
F05 - Treated - 96

When doing variant calling, the normal tissue is specified with an n and the tumor tissue with a t. For example, the output file of n98_t96 would be the 96 sample (F05) is the t = tumor and 98 sample (Control) is n = normal when doing variant calling.

Important

Contact