Drug Discovery & Cheminformatics

Cheminformatics Scientist

Apply RDKit, docking, and generative chemistry to accelerate small-molecule drug discovery.

Build my Cheminformatician roadmap Compare salaries See open roles

Cheminformaticians sit inside medicinal chemistry teams, applying computational tools to triage millions of compounds, predict ADMET liabilities, and generate novel chemical matter. The role bridges chemistry, ML, and structural biology.

Cheminformatics Scientist salary (USD)

entry

$90k–$120k

mid

$120k–$170k

senior

$170k–$260k

US base ranges blended from Levels.fyi, BLS, Glassdoor, and Payscale (2024–2025). See full salary benchmark →

What a Cheminformatician does day-to-day

Maintain compound libraries and search/filtering pipelines.
Run docking and free-energy calculations on hits.
Build QSAR and ADMET models on internal and public data.
Apply generative chemistry to expand chemical series.

Required skills & tools

Core knowledge

Medicinal chemistry awarenessMLMolecular dynamics basicsCheminformatics QC

Tools

RDKitAutoDockSchrödingerOpenEyeDeepChemChemprop

Languages

PythonC++

12-month roadmap to Cheminformatician

1
RDKit (0–3 mo)
SMILES, descriptors, fingerprints, similarity searches.
2
Modeling (3–6 mo)
QSAR with scikit-learn, ADMET prediction, model validation.
3
Structures (6–9 mo)
Docking with AutoDock/Vina, MD basics with GROMACS.
4
Generative (9–12 mo)
REINVENT / Chemprop / diffusion models on public targets.

Personalize this roadmap with AI

Job titles to target

• Cheminformatics Scientist
• Senior Cheminformatician
• Principal Scientist, Computational Chemistry

Where they hire

• Big pharma
• Biotech drug discovery
• CROs
• AI-first drug discovery startups

FAQ

Do I need a chemistry PhD for cheminformatics?

Senior roles in big pharma typically expect a PhD in chemistry, computational chemistry, or a closely related field. Engineering-leaning and startup roles are accessible with an MSc plus strong RDKit and ML portfolio.

Is generative chemistry replacing medicinal chemists?

No — it augments them. Generative models propose candidates; medicinal chemists triage, synthesize, and test. Cheminformaticians sit between the two.

What public datasets should I learn on?

ChEMBL for bioactivity, PubChem for general structures, BindingDB for affinities, and the MoleculeNet benchmarks for ML model training.