Biotechnology engineering student (ESBS, University of Strasbourg; MSc Biology, Freiburg) moving into computational biology: modelling, molecular docking/dynamics, structural bioinformatics and AI/ML in biology.
Looking for a 6-month end-of-studies internship starting February/March 2027 (degree expected August 2027).
📫 [daniel.ogundero@etu.unistra.fr] · LinkedIn
Question: how does a nylon-degrading enzyme (NylC, PDB 3AXG) bind a polyamide substrate, and does docking into the native tetramer change the outcome compared with the isolated monomer? Relevant to enzymatic plastic recycling.
What I did: built a reproducible Python pipeline around AutoDock Vina, MGLTools and PyMOL (receptor/ligand preparation, docking, result parsing, 5-seed replicates). Poses were ranked by score and by catalytic geometry (distance between Thr-267 OG1 and the ligand's carbonyl carbon).
Findings: Vina’s score alone did not identify the reaction-plausible pose, so geometry was checked in both conditions. The tetramer scored about 0.5 kcal/mol better than the monomer across replicates: reproducible, but below typical Vina scoring uncertainty, so not strong evidence of a real binding difference. This is tentatively attributed to atoms from other chains falling inside the search box (hypothesis, not yet tested with a control).
Skills: structure preparation, docking, pose analysis, scripting, honest reporting of limitations (rigid receptor, score ≠ affinity). Built independently, with AI assistance for learning and troubleshooting, as disclosed in the repository.
Question: how does removing two binding-pocket aromatic residues change the dynamics of BRD4 BD1 and the stability of a bound dual BET–HDAC inhibitor (LSH24)?
What I did: compared apo, holo and mutant systems (system building with CHARMM-GUI, CHARMM36m/CGenFF, NAMD, analysis in VMD and PyMOL). Coursework at ESBS, done with a classmate: we each ran the full set of MD simulations independently on our own machines, then co-wrote the report together.
Findings: a short single-run vacuum simulation (200 ps), so conclusions are limited: the ligand stayed in the pocket in both holo and mutant, and the mutant showed a late localised rearrangement that cannot be attributed to the mutation from one run. The repository states these limitations explicitly.
Skills: MD setup and minimisation, force-field parametrisation, trajectory analysis, critical interpretation.
Python · AutoDock Vina · MGLTools · PyMOL · NAMD · VMD · CHARMM-GUI · CHARMM36m / CGenFF · Git/GitHub
French and English (bilingual) · Spanish (B1) · German (A2)
Structural bioinformatics · molecular modelling and dynamics · AI/ML applied to biology · high-throughput biological data analysis