This J Med Chem editorial (8 October 2026) presents direct-to-biology (D2B) platforms: crude or minimally processed reaction mixtures are assayed without purification, collapsing the cycle into “synthesis, then biological test”. The gain is clearest for bifunctional molecules (PROTACs, molecular glues), whose multi-step syntheses limit library diversity.
Examples cited: a GSK D2B-HTC platform generated more than 1000 compounds in 384-well plates in one day; the Cernak group reached a 73% success rate in 1536-well format using 0.03 mg of natural product per reaction; for MNK inhibitors, more than 150 analogues were tested directly in cells, with IC50 values as low as 23 nM; for Cdk2/CycE, 768 compounds were made by amide coupling of 128 acids to six aminopyrazole cores; in targeted degradation, SuFEx chemistry installed 3163 building blocks on ENL and BRD4 ligands, and a photoclick platform produced more than 1000 cereblon-targeting molecular glues in plates.
The authors note that data from crude mixtures agree strongly with measurements on isolated, purified compounds. The limits are real, though: matrix interference (catalysts, by-products) can skew fluorescence or luminescence readouts, heavy metals can disrupt target proteins, and the usable reaction set is narrow (SNAr, amide, CuAAC, SuFEx). Mitigations discussed include matrix controls, metal scavengers and orthogonal label-free validation such as SPR. Outlook: wider reaction scope and AI-assisted library design.
Takeaway: D2B is ideal for exploring SAR quickly on click-type chemistry; any hit from a crude mixture should be confirmed on purified compound and with an orthogonal biophysical method.
Reference: 10.1021/acs.jmedchem.6c02736, J Med Chem 2026, 69, 21539–21542.