Pharmacoinformatics & Drug Design

DrugBank is a pharmaceutical knowledge database and a key pharmacoinformatics resource with comprehensive drug/target/disease information. In this Jupyter notebook, we will download approved drugs information for Alzheimer´s Disease from DrugBank. More information here.

Have fun!!!


Here we use a KNIME workflow to post-process the output file *CSV* from a multiple pharmacophore-based virtual screening calculation done with Phase. We screening a dataset of ~1000 compounds against multiple pharmacophores (~100), then the results were exported as a *CSV* file (see [*/Files/screening_phase_default-otro.csv*]( with all the results with the corresponding phase screen score (0 to 1). This workflow organizes all the results and creates a matrix (compounds vs. pharmacophores) as well as a heatmap figure. More information here.

The workflow was programed with KNIME v 4.1.0, and can be found here.


Open Target Platform

PSC-db is a free accessible database containing the 3D-structures of several Plant Secondary Compounds along with its physicochemical and pharmaceutical properties.


Using AutoDock

We have designed a pipeline to analyze in detail all interactions of one or several lioagnd-receptor complexes. For this we use native scripts from the Schöridnger Suite and the data science software KNIME. In this way we calculate an interaction profile of a complexes cluster that migth come, i. e., from massive docking simulations or a trajectory from molecular dynamics simulactions. More information here

Here we use AutoDock Vina to perform massive docking of one ligand: 100 runs -> top-10 poses per run -> 1000 poses. The idea is to use a single ligand and do multiple dockings on the same receptor to extend the conformational sampling of that ligand. Then you can use a tool to cluster the conformers and know which are the most visited poses during the molecular docking. More information here.

Here we use Glide (Schrödinger Suite) to perform massive docking of one ligand into several conformations of the same receptor extracted from a molecular dynamics simulation. The idea is to include the flexibility of the receptor binding site (side chains) into a large conformational sampling. More information here.

Using LigandScout

Here we use a tcl script to count the water molecules in an ion channel pore (or any pore) along a trajectory from AMBER, Desmond or NAMD. The pore of the ion channel must be aligned along the z-axis. The selection must be a cylinder.. More information here.

Using Gromacs and ff-Martini --> Ionic Liquis

With Ligand Scout

Here we explain how to properly load and visualize Desmond's trajectories in VMD.

  1. Open VMD and load a desmond_trajectory-out.cms file as new molecule.
  2. Load the clickme.dtr trajectory file (it is located into the desmond_trajectory_trj folder) into the desmond_trajectory-out.cms file.
  3. To fix the periodix boundary conditions, you need to center the trajectory to generate a continuous trajectory view with the protein centered in the primary simulation box, to do that use the loaded trajectory and try either of the following commands from VMD’s Tk console (Extensions > Tk Console):
    • pbc wrap -centersel protein -center com -compound res -all
    • pbc wrap -center bb -centersel protein -sel “not protein” -compound res -all
  4. You may have to superimpose all frames of the trajectory onto a selected frame in order to center the view in the main VMD graphics window. To do so, select VMD > Extensions > Analysis > RMSD Trajectory Tool, and select the Align option. Note that the transformed trajectory can be written to disk for further analysis.

Have fun!!!

A complete solution for fast binding protein-ligand based on molecular dynamics using BIKI

Here we use a tcl script to get the resdiues within a given distance of a ligand during a trajectory. You can use it to study Protein-Protein or Peptide-Protein interaction as well. More information here.

TASK-1 + A1899

Using Modeller and Rosetta...

pasos a seguir...

HOLE is a program that allows the analysis and visualisation of pore dimensions of the holes through molecular structures of different proteins such as ion channels. Here we will use HOLE suite of programs to study pores in ion channels. More information here.

Some examples:



1 small GPU cluster (Tensor TWS-1686525-AMB: 2x Intel Xeon Silver 4110 Processor, 8 Core, 16 Threads, 2.1Ghz, 11M Cache, 85W. 12x 8GB DDR4 ECC/REG 2666MHZ (96GB System memory). 1x 256GB 2.5 SATA III Internal Solid State Drive (SSD). 1x 8TB SATA 6Gb/s 7200RPM 3.5" 256MB Cache Hard Drive. 3x GPU Nvidia Quadro P5000 16GB. CentOS 7)

4 workstations for visualization and data analysis (1x Intel CPU Core i7-7700 3.6GHz. 1x Toshiba Disco Duro 2TB Sata3 7200 rpm 64MB P300. 1x SSD 480GB Sata3 2.5". 2x MSI Video Nvidia Gaming GeForce GTX1660 6GB. 2x Kingston DDR4 16GB 2400MHz HyperX FURY Black).

We have access to the CPU- GPU-clusters at the Center for Bioinformatics and Molecular Simulations at Universidad de Talca, Chile


Software for molecular simulation, modeling and Visualization: Schrödinger Drug Design Suite, Modeller, AMBER, GROMACS, PyMol, VMD, KNIME, AutoDock, AutoDock vina, LigandScout, BIKI.


The Facultad de Ciencias Biológicas at the Universidad de Concepción is integrated by the departments of cell biology, biochemistry, ergonomics, pharmacology, physiology, physiopathology, and microbiology. Each department has different equipment and facilities that allow them to do cutting-edge research in their area of expertise.