Endogenous Deep Mutational Scans (EDMS)

Command Line Interface

edms -h # or edms <TAB>

Library Design Pipelines

Two commands chain a whole library design together instead of running each step by hand. Both take --query (pandas query strings, applied in order) and --drop_duplicates to filter the designer output, and --signature_query / --signature_drop_duplicates to filter again on the signature columns (difference_count, SNV_count, ...).

edms pe pipeline — pooled epegRNA libraries

prime_designer -> filter -> pegRNA_signature -> filter -> epegRNA_linkers -> epegRNA_pool

epegRNA_linkers is the slow step, so it is fanned out across --linker_jobs SLURM scripts rather than run inline. That splits the command into two stages:

# 1. Design, filter, and write out_dir/linkers/{chunk_N.sh, pool.sh, submit.sh}
edms pe pipeline --stage design -n FOXA1 -f5 ... -t ... -f3 ... -x 168 \
    -sm aa_silent -plp 11 -npe 20 -nng 1 \
    -q "Reference_sequence.str.contains('ACGTTCAAGCGCAGCTACCCG')" \
    -dd Edit -C spMUZ88_spMUZ89 -sq "difference_count == 1" \
    -j 10 -bval "FOXA1[DBD]" -o ./pe_pipeline

# 2. On the cluster: submits every chunk, then chains --stage pool behind them (afterok)
cd pe_pipeline/linkers && bash submit.sh

--stage pool reads its settings back from out_dir/.pe_pipeline.json, so nothing has to be retyped; --stage all runs everything in one process with serial linkers (small libraries only).

edms dms pipeline — pooled DMS libraries

dms_designer -> filter -> dms_signature -> filter -> dms_pool, all in one process.

edms dms pipeline -n FOXA1 -f5 ... -t ... -f3 ... -x 168 \
    -sm aa -s 2 -smode barcode -C spMUZ88_spMUZ89 \
    -q "Edit_type == 'substitution'" -dd Edit -sq "difference_count == 3" \
    -fht5v FOXA1-S165 -rht3v FOXA1-P205 -o ./dms_pipeline

dms_pool is skipped (with the command to run later) unless both homology values are given.

Package Organization

• gen: input/output, data wrangling, generating plots, and statistics.

  import edms.gen.com as com
  import edms.gen.html as ht
  import edms.gen.image as im
  import edms.gen.io as io
  import edms.gen.plot as p
  import edms.gen.stat as st
  import edms.gen.tidy as t
  

• bio: molecular biology & tissue culture workflows.

  import edms.bio.clone as cl
  import edms.bio.dms as dms
  import edms.bio.fastq as fq
  import edms.bio.genbank as gb
  import edms.bio.ngs as ngs
  import edms.bio.pe as pe
  import edms.bio.pegLIT as pegLIT
  import edms.bio.primedesign as primedesign
  import edms.bio.qPCR as qPCR
  import edms.bio.sanger as sanger
  import edms.bio.signature as signature
  import edms.bio.transfect as tf
  

• data: interacting with databases.

  import edms.data.cBioPortal as cBP
  import edms.data.cosmic as co
  import edms.data.cvar as cv
  import edms.data.pdb as pdb
  import edms.data.uniprot as uniprot
  

PyPI Instructions

Install

1. Install edms from PyPI

   pip install edms
   

2. Check edms install

   edms -h
   

3. Optional: Set up edms autocomplete

   edms autocomplete
   # Follow CLI instructions
   

4. Optional: edms fastq {extract_umis,trim_motifs,make_sams,make_bams,bam_umi_tags,group_umis,consensus_umis,bam_to_fastq} need umi_tools, cutadapt, samtools, bowtie2, and fgbio in a seperate environment

   conda create -n umi_tools umi_tools cutadapt samtools bowtie2 fgbio
   

Update

1. Update edms from PyPI

   pip install --upgrade edms
   

GitHub Instructions

Install

1. Download Anaconda:
   • Mac: https://docs.anaconda.com/anaconda/install/mac-os/
   • Windows: https://docs.anaconda.com/anaconda/install/windows/
   • Linux: https://docs.anaconda.com/anaconda/install/linux/
2. Download Git: https://github.com/git-guides/install-git
3. Clone edms from github:

   cd ~
   mkdir git
   cd git
   git clone https://github.com/marczepeda/edms.git
   cd edms 
   

4. Make the environment and install edms:

   conda env create -f edms.yml # When conda asks you to proceed, type "y"
   conda activate edms
   pip install -e . # Include the "."
   edms autocomplete # Optional: set up edms autocomplete; follow CLI instructions
   conda deactivate
   

5. Optional: edms fastq {extract_umis,trim_motifs,make_sams,make_bams,bam_umi_tags,group_umis,consensus_umis,bam_to_fastq} need umi_tools, cutadapt, samtools, bowtie2, and fgbio in a seperate environment

   conda create -n umi_tools umi_tools cutadapt samtools bowtie2 fgbio
   

Update

1. Enter the environment and uninstall edms:

   cd ~/git/edms
   conda activate edms
   pip uninstall -y edms
   rm -rf build/ dist/ *.egg-info
   

2. Pull latest version from github and install edms:

   git pull origin main
   pip install -e . # Include the "."
   conda deactivate
   

PE Strategies