Skip to content

Roadmap to version 1.0.0

Diego Prada-Gracia edited this page May 13, 2021 · 1 revision

Towards version 1.0.0

Let's see if this can be a way to agree on a set of features and goals to be achieved before reaching version 1.0.0. Once we have a list, let's try to sort them in 99 steps (maximum). This way we will be defining a tentative roadmap from version 0.0.0 to 1.0.0.

Items, concepts, features, goals, we want to see in OpenPharmacophore 1.0.0

According to the semantic versioning rules, the major version jumps when the software evolves in such a way that the continuity in the API use is broken. But we want to avoid this no matter what new direction OpenPharmacophore is going to take. So, let's set from now on a more conceptual criterium. If you all agree, version 1.0.0 builds pharmacophores from ligand-based and structure-based (and ligand-structure-based) cases taking common and validated strategies from the literature. And it does so with structures coming from molecular dynamics in a flexible and modular way to be able to perform virtual screenings from opened databases. Below, let's break down all features and concepts we want to implement before reaching 1.0.0. Once the list is close to drawing a complete roadmap, it will be sorted in discrete increments to define the minor version numbers.

  • Pharmacophoric points definition from a ligand-based analysis.
  • Visualization of pharmacophoric points in NGLView.
  • Pharmacophore model definition (agnostic -no matter how it was created-).
  • Pharmacophore models trajectories from molecular dynamics.
  • Working with pharmacophore models to get consensus models or "representative" models.
  • Markov State Models of pharmacophores to determine the most interesting models to be used for screening
  • Screening chemical compounds with a pharmacophore model.
  • Something about validation of the workflow?
  • Something about confidence estimation of the results?
  • Enhance sampling strategies, such as adaptive sampling, to harvest the statistically relevant models avoiding brute force simulations.
  • First CI strategies implementation.

Dreams for OpenPharmacophore version > 1.0.0

  • Pharmacophore models from simulations with chemical probes.