Software
We are proud to develop and contribute the following open-source tools to the broader polymer community. For each tool, we include a brief description, relevant publication, and link to the source code for installation. Please visit our GitHub page for more information.
Percent Collapse Analysis and Copolymer Solution Conformation
ML (Copolymer Solution Conformation): Python workflows for polymer machine learning and inverse design, including feature and encoding calculations, model fitting and evaluation, feature ablation, and genetic algorithm-based optimization.
SEC (Percent Collapse Analysis): Python notebooks for processing LabSolutions size-exclusion chromatography exports, generating molecular weight chromatograms, and calculating polymer percent collapse.
PRISM
Polymer rate insights and sequence modeling: Enabling reconstruction and analysis of synthetic polymer sequences. Prism includes modules to estimate monomer reactivity ratios from kinetics data of RAFT polymerization from two- and three-monomer systems, a kinetic stochastic simulation strategy to reconstruct these sequences, and a variety of metrics to enable visualization and quantification of chemical patterning.
Particle-Counting
Extracting population data from electron and atomic force microscopy data is challenging and time intensive. Particle-Counting automates the initial stages of microscopy data processing by identifying all spherical particles in a given acquisition and measures them. This tool interfaces with the free ImageJ software (https://imagej.net/ij/).
PeptoidSeq
Peptidomimetics remain challenging to sequence due to noncanonical monomers and linkages. PeptoidSeq facilitates rapid sequencing of peptoid (N-substituted glycines) by reading a MALDI TOF-TOF spectrum and outputting a candidate sequence. This tool also supports user input to resolve ambiguous assignments. PeptoidSeq was developed and validated on a binary monomer peptoid system consisting of glycine and N-butylglycine.
ColorClassify
One-bead one-compound (OBOC) libraries are high-throughput strategies for synthesis and characterization of macromolecules. ColorClassify quantifies images of OBOC libraries stained for a colorimetric assay. It consists of two parts: (1) an image quantification code that returns RGB values of selected beads in an image, and (2) a machine-learning classification algorithm that will classify lists of RGB values into groups based on pre-classified test data.
MotifFold
MotifFold is an implementation of a data-driven strategy towards elucidating sequence-structure relationships for synthetic macromolecules. It uses a language-inspired "motif" strategy, where sequences are separated into motif features and then regressed against a conformation descriptor. MotifFold was developed to describe sequence-dependent folding in peptoids as quantified by a colorimetric screen.
Copolymer Sequence Generation
Synthetic polymers are stochastic objects with dispersities in sequence and length. CopolymerSequenceGeneration is a tool to generate an ensemble of polymer sequences given the reactivity ratios of the composite monomers. Using a Kinetic Monte Carlo strategy, it simulates the chain growth and monomer incorporation process, and comparisons between various ensembles can be performed through calculation of Jensen-Shannon divergence.