ACE Configurator for ELISpot

ACE facilitates (1) generation of ELISpot configurations (peptide-pool assignments) using a deep learning approach to cluster similar peptides and (2) deconvolution of pool spot counts for identification of immunogenic peptides.

01. Installation

01-1. Standalone Graphical User Interface (Recommended)

Please note that the ACE GUI software will take a long time to load (~30 seconds ☕). We also recommend that you have Google Chrome installed on your machine.

Operating System	Link	Version
Mac	Download	v0.1.1.0 (latest)
Windows 10	Download	v0.1.1.0 (latest)

For Windows versions, first unzip the file and look for an application file called ACE inside the unzipped foler. Previous versions of ACE are available here.

01-2. Python Package

ACE is available on PyPI

pip install ace-elispot

You can also download a specific version of ACE from here.
Subsequently install the ACE package using pip:

pip install ace-elispot-<version>.tar.gz

Dependencies

python3 (>= 3.10)
pandas (>=1.5.2)
numpy (>=1.23.1)
ortools (9.3.10497)
torch
transformers (==4.30.2)
scikit-learn
openpyxl
golfy (>=2.5.0)
levenshtein

02. Usage

ACE is available as a command-line interface after you install the python package:

usage: ace [-h] [--version] {generate,deconvolve,verify} ...

ACE Configurator for ELISpot.

positional arguments:
  {generate,deconvolve,verify}
                        ACE sub-commands.
    generate            Generates an ELISpot experiment configuration.
    deconvolve          Deconvolve hit peptide IDs given read-outs from an ELISpot experiment.
    verify              Verifies whether an ELISpot assignment satisfies all ACE constraints.

options:
  -h, --help            show this help message and exit
  -v, --version         show program version number and exit

Read the full documentation on the python package at https://pirl-unc.github.io/ace/

03. Citation

If you use ACE in a publication, please cite our publication describing ACE.

dhuvik/ace