evSeq

No sequence-function pair left behind.

Every Variant Sequencing (evSeq) is a library preparation and analysis protocol that slots neatly into existing workflows to enable extremely low-cost, massively parallel sequencing of protein variants. Designed for heterologously expressed protein variants arrayed in 96-well plates (or similar), this workflow enables sequencing all variants produced during a protein engineering or biochemical mutagenesis experiment at a cost of cents per variant, even for labs that do not have expertise in or access to next-generation sequencing (NGS) technology.

Read the Paper!

This repository accompanies the work "evSeq: Cost-Effective Amplicon Sequencing of Every Variant in Protein Mutant Libraries".

Read the Docs!

For detailed information and interactive walkthroughs, read the docs at the evSeq website.

Quick links to common resources:

Biology

• Inner Primer Design
• PCR Protocol
• PCR Product Purification

Computation

• Installation
• Inputs
  • The refseq file
• Outputs
  • OutputCounts
• Running evSeq in a Jupyter Notebook

Troubleshooting

The evSeq workflow

A) evSeq amplifies out a region of interest that contains variability, attaches well-specific barcodes and adapters, and is ready for NGS.

B) All that's required to perform the evSeq laboratory procedure is:

• a 96-well thermalcycler
• standard PCR reagents and materials
• access to an NGS provider
• two 96-well plates of evSeq barcoding ("outer") primers
• a pair of region-specific evSeq-compatible ("inner") primers
• 96-well plate(s) of cultures containing DNA encoding protein variants
• a 12-channel 10 µL pipette is also helpful; robotic support could be developed to replace many manual pipetting steps

That's it.

Due to the two-primer, culture-based PCR methodology employed by evSeq, only a new pair of inner primers needs to be ordered when targeting new regions/sequences and no DNA isolation needs to be performed.

C) Once the sequences are returned by the NGS provider, the computational workup can be performed on a standard laptop by users with little-to-no computational experience.

The amplicons prepared with evSeq can yield nearly 1000 high-quality protein variant sequences for just the cost of the multiplexed NGS run (typically ~$100 from commercial sequencing providers and much less from in-house providers).

Construct and visualize sequence-function pairs

Sequencing eight site-saturation libraries (768 wells) in a single evSeq run and combining this with activity data to create low-cost sequence-function data. A) Enzyme and active-site structure highlighting mutated residues. B) Heatmap of the number of identified variants/mutations ("counts") for each position mutated ("library") from processed evSeq data. C) Heatmap of the average activity ("normalized rate") for each variant/mutation in each library. D) Counts for a single library, also showing the number of unidentified wells. E) Activity for a single library, showing biological replicates. (Inset displays the mutated residue in this library.)

Installation

RECOMMENDED. Use the evSeq environment:

git clone https://github.com/fhalab/evSeq.git
cd evSeq
conda env create -f envs/evSeq.yml

evSeq is then installed inside the environment and can be run as described below when the evSeq environment is active.

This also installs a shortcut to the GUI on your Desktop (which can then be moved, e.g., to an Applications folder) which will run evSeq in the proper environment simply with a double click (see below).

Alternatively, you may install evSeq directly from PyPI via:

pip install evSeq

The correct packages should be automatically installed and a GUI shortcut is also made, but this is not guaranteed to work as you update your packages/dependencies in the future.

Updating

By default, evSeq is installed in non-dev mode. This means that changes to the code base on your computer will not be reflected come run-time. If you want an editable version of evSeq, install with the evSeq_dev environment (note, however, that this environment does not set exact versions of dependencies like the evSeq environment). We recommend installing in non-dev mode (i.e., using the evSeq environment). Finally, we also provide the evSeq_general environment file which is similar to evSeq_dev but does not install an editable version of evSeq. To update evSeq when installed in a non-dev environment, the environment must be recreated. The below commands will update evSeq. First, navigate the evSeq repository folder via command line and enter the below commands:

git pull
conda remove -n evSeq --all
conda env create -f envs/evSeq.yml

Usage

Command Line

Thanks to setuptools entry_points, evSeq can be accessed from the command line after installation as if it were added to PATH by running:

conda activate evSeq
evSeq refseq folder --OPTIONAL_ARGS ARG_VALUE --FLAGS

where refseq is the .csv file containing information about the experiment described above and folder is the directory that contains the raw .fastq files (.gz or unzipped) for the experiment.

For information on optional arguments and flags, run

evSeq -h

or visit the usage page.

GUI

evSeq is also installed with a GUI for greater accessibility. After installing evSeq, you will find a new shortcut/app/executable on your Desktop, which you can double-click to launch the evSeq GUI:

More can be read about it here.