What is MolinTime?

An end-to-end in silico hit discovery service that delivers a ranked shortlist of compounds, tailored to your target and screened for predicted binding affinity, ADMET profile and novelty, while remaining within synthetically accessible chemical space.

Hit discovery on your target

Our hit discovery service starts from your target and screens vast variety of the chemical spaces

How we work

Hit Identification

We screen your chosen vast chemical space with a multiparametric pipeline and return a ranked initial shortlist

Structure-based methods to pose and score compounds against the binding site, and evaluate stability & selectivity

Ligand-based methods for affinity & selectivity prediction

Novelty assessment against patents, scientific literature & public databases

In silico ADMET evaluation for solubility, permeability, metabolic stability & toxicity

Hit Expansion

  • From confirmed hits, we generate new analogues using reaction-based methods and Matched Molecular Pairs (MMP)-based approaches
  • Every generated molecule is restricted to synthetically accessible chemical space and scored with the same multiparametric pipeline used for hit identification

What you receive

A ranked list of compounds, prioritised for synthesis and biological testing

Case Study

Pikaïros computational hit design paired with Enamine synthesis & experimental validation.

Target implicated in inflammatory, autoimmune, and oncological disease.

Hit Identification

Starting from Enamine’s in-stock collection (~4 M compounds)

~4 M

compounds in the screened collection

160

compounds were computationally preselected and experimentally assayed

1

confirmed hit at 230 nM (0.23 µM)

Hit Expansion

We generated new analogues within Enamine REAL Space and scored them using the same pipeline:

174
compounds tested
71
with >50% inhibition at 20 µM
45
with >70% inhibition at 20 µM
40
compounds selected for dose-response
16
of 40 confirmed as submicromolar

40 compounds selected for dose-response

Best lead

35 nM

next best 81 nM

  • IC₅₀ range 35 nM – 5.3 µM (median ~1.2 µM)
  • 16 of 40 confirmed as submicromolar
  • Best lead 35 nM, next best 81 nM

Across the 174 selected molecules, mean nearest-neighbour ECFP4 Tanimoto similarity was 0.38 to ChEMBL 37 and 0.35 to target-associated compounds drawn from public databases and patents

Network of 40 compounds coloured by IC₅₀: 2 below 100 nM, 14 between 100 nM and 1 µM, and 24 between 1 and 5.3 µM. Lines connect pairs with ECFP4 Tanimoto similarity of 0.50 or more; most compounds fall into several small clusters or remain singletons.
40 dose-response confirmed hits, edges denote similarity* ≥ 0.50. *ECFP4 Tanimoto, 2048-bit

Profile of Selected Molecules (n = 174)

Nearest-neighbour similarity* to ChEMBL 37

Mean 0.38 · Max 0.51

Nearest-neighbour similarity* to target-associated compounds (full molecule)

Mean 0.35 · Max 0.46

Nearest-neighbour similarity* to target-associated compounds (scaffold)

Mean 0.38 · Max 0.55

Structural diversity

146 of 174 unique Murcko scaffolds

*ECFP4 Tanimoto, 2048-bit

Tailored Chemical Collections

Each library is built around a single drug target and delivers a ranked, annotated set of molecules, from hundreds to thousands per collection, generated through machine learning and docking. Every molecule is sourced from commercial catalogues, with predicted potency, selectivity, and ADMET profile already assessed, so it is ready for synthesis and experimental screening without further preprocessing.

TargetUniProtTherapeutic areaDesign ObjectivesCollection size In Silico Design methodIn silico ADMET profiledStatusYear
Tyrosine-protein kinase Mer (MERTK)Q12866Oncology, ImmunologyType I, ATP-competitive, TAM-family aware design1105 moleculesML, DockingYesAvailable2025
Cyclin-dependent kinase 9 (CDK9)P50750Oncology, CardiovascularATP-competitive, CMGC-family aware design915 moleculesML, DockingYesAvailable2026
Tyrosine-protein kinase BTK (BTK)Q06187Oncology, AutoimmunityNon-covalent, ATP-competitive450 moleculesML, DockingYesAvailable2026
Poly [ADP-ribose] polymerase 1 (PARP1)P09874 OncologyNicotinamide-site directed850 moleculesML, DockingYesAvailable2026
Phosphatidylinositol 4,5-bisphosphate 3-kinase catalytic subunit alpha isoform (PI3K-alfa)P42336Oncology ATP-competitive, class I PI3K910 moleculesML, DockingYesAvailable2026
Dihydroorotate dehydrogenase (quinone), mitochondrial (DHODH)Q02127Oncology, Autoimmunity Ubiquinone-tunnel directed758 moleculesML, DockingYesAvailable2026

Browse our available libraries or contact us to discuss licensing, series selection, or a custom collection for your target.

How were collections designed?

01

Target input & biological rationale

Target biology, disease context, known chemical space, and available structural data are assessed to define a concrete set of hit-identification objectives.

02

Database & patent curation

Public chemical and biological databases are mined alongside patent landscapes to map existing chemical matter. Novelty is benchmarked against this auditable reference set, ensuring each delivered molecule is potentially novel relative to a defined baseline.

03

Chemical space exploration

Novel candidates are generated through ML-driven ligand-based and structure-based methods to identify molecules with strong potential to be active against the target of interest.

04

Multi-criteria prioritization

Candidates are scored across novelty, predicted potency, ADMET profile, and synthetic accessibility. Compounds flagged for hERG inhibition, structural alerts, PAINS, or target-specific liabilities are excluded. Models are validated against held-out experimental data.

05

Output

A ranked collection of hundreds to thousands of structurally diverse molecules, each sourced from commercial vendor catalogues with supplier name, indicative pricing, and lead time. Here, synthetic accessibility means a compound ready to order and assay.