Edit on GitHub

First steps

Introduction

The rational placement of amino acids in a sequence directly correlates to the 3D structure of the peptide once folded. This folded structure, in turn, dictates the interactions with the environment and therefore the function of the peptide. Being able to design new peptides for specific functions would unlock the potential of unknown conformations not yet explored by nature, with applications in medical fields, agriculture, biological remediation, enzymatic synthesis, among others.

This was traditionally explored by random blind mutagenesis which is an expensive and time intensive experimental practice. With the evolution of computational power over the last couple of decades, computational design of small proteins has become the focus of scientific breakthroughs. Using computer simulations saves precious time and monetary costs of experiments, focusing efforts on simulated prototypes that have shown promising results. As such, multiple software solutions have been proposed over the years. An example would be Rosetta (and its Python wrapper, PyRosetta), which has been indisputably invaluable as a platform for molecular manipulation and peptide design.

However, as it happens with so many scientific-purposed software packages, Rosetta has fallen into the two-language problem. The core of the simulation code is written in C with a more user-friendly wrap in Python that exposes some of the functionality. This has gravelly impaired the community's ability to upgrade and modernize this package over the years, as well as imposing a steep learning curve to non-specialized would-be-users.

ProtoSyn.jl, empowered by the Julia language ecosystem, aims to put forward a simple, easy to use and completly open-source framework for molecular manipulation and simulation, with a focus on peptide design. A Julia-based solution to this challenge would naturally benefit from the native features of the language, such as easy parallelization and distributed computing, GPU acceleration, machine learning tools, and a growing community of enthusiastic scientists, among others.

Under the hood: how is ProtoSyn organized?

organization

ProtoSyn.jl's main struct is the Pose, which holds all the required information regarding a molecular system. This information is divided into a directional graph and a state.

A directional graph is simply an hierarchical structuration of the molecular system: At the top level, a Topology can hold multiple Segment instances (which are contiguous chains of a molecule), which in turn can hold multiple Residue instances (such as amino acids, in the case of proteins and peptides), which, finally, hold one or multiple Atom instances.

Each of these levels are called AbstractContainers and can have defining parameters. In the case of atoms, for example, an Atom instance is described by an :id, :symbol, :name, etc.

This graph is called directional because each of these components has a parent and can have one or more children containers. In the case of Residue instances, for example, consider the sequence ALA-GLC-PRO. In such a molecule, ALA would be the parent of GLC which, in turn, would be the parent of PRO. The same logic applies to relationships between atoms. This structural organization allows ProtoSyn to easily travel the Graph of a structure during certain molecular manipulation processes. Perhaps more interestingly, when conjugated with the internal coordinates system, this data organization system allows for easy and fast manipulation of dihedral angles, bonds distances, etc.

The information regarding these internal coordinates (and cartesian coordinates) is organized in the pose's State. This is subdivided into a list of AtomState instances and a StateMatrix. Both of these structures are interchangable, meaning that altering a value in StateMatrix updates the corresponding value in the correct AtomState and vice-versa. The rational behind having both structures lies in having the ease/speed of changing a large volume of coordinates in the StateMatrix at once, while still being able to control internal coordinates in single atoms using the AtomState. In sum: some manipulation processes benefict from using cartesian coordinates, while others are either faster or simpler using the internal coordiantes.

In order to initialize the internal coordinate system, each Topology has an extra set of 3 pseudo atoms, called a Root, that sets-up the ascendents for the first few atoms of the molecule.

Mutators, Energy Functions & Drivers

organization

ProtoSyn.jl offers some quick simulation functionalities, such as MonteCarlo, ILS and SteepestDescent simulations, using Driver instances (see Drivers section). A Driver is a function which drives the pose from one state to the next. As a general rule, these Drivers are usually comprised of two important components: one or more Mutator instances (see Mutators section) and an Energy Function (see Calculators section).

A Mutator is a function that performs a single change in the system. For example, a DihedralMutator will rotate a random dihedral by a random amount. These can be parametrized for more specific needs.

An EnergyFunction evaluates the fitness of a given Pose based on a set of EnergyFunctionComponent instances. These can be distance-based restrictions, positional agreement with machine learning models (such as TorchANI), etc.

Note:

Some commonly used EnergyFunction instances can be found in the Common module.

Modular system

ProtoSyn.jl package is organized in a modular fashion, with each new module adding specific methods and features relative to a given scientific area of expertise. All these modules build on top of the base Core module. For example, in the Core module, ProtoSyn.load is able to read a PDB file into a Pose struct, but calling ProtoSyn.Peptides.load (from the Peptides module) will also infer peptidic connections between amino acids, setting the correct parenthood relationships between residues in the pose's graph and rename Atom instances to the expected IUPAC convention used in peptides, returning a much more information complete Pose.

Next steps

This section aimed to provide a very brief and general overview on how ProtoSyn is organized, but there so much more to learn and explore: