Liganx blog

Drug targets, resistance mutations, ADMET properties, and molecular docking methodology — written for researchers in structure-based drug discovery.

• CDK4/6 inhibitors in HR+ breast cancer: three approvals, one class

  Palbociclib, ribociclib, and abemaciclib all hit the same kinase pair but differ on selectivity, dosing, and toxicity. Here is how the class actually splits.

• DFG-out and Type II kinase inhibitors: where selectivity comes from

  Imatinib, sorafenib, ponatinib, nilotinib — Type II inhibitors all bind a conformation the ATP-mimetic Type I drugs never see. Here is what DFG-out is and why it matters.

• Ensemble docking: when one rigid receptor isn't enough

  Why docking against a single crystal structure misses real binders, and how docking across multiple receptor conformations recovers the ones a rigid pocket throws away.

• KRAS codon 61: why Q61 mutations break the GTPase

  Codon 61 mutations disable the catalytic glutamine that hydrolyzes GTP, locking KRAS in the ON state. What that means for SHP2 resistance and pan-RAS drugs.

• Aqueous solubility: the property that quietly kills oral programs

  Kinetic vs thermodynamic solubility, the BCS map, and why a beautiful nanomolar binder can still fail because it will not dissolve. A practical explainer.

• EGFR L858R: the activating mutation that built a drug class

  L858R is one of the two classic EGFR drivers in lung cancer. Here is how a single residue swap in the activation loop turns a kinase on, and which drugs shut it back off.

• ALK L1196M: the gatekeeper mutation and the lorlatinib answer

  L1196M was the first crizotinib-resistant ALK mutation found. Here is the gatekeeper mechanism, the drugs that overcome it, and why compound mutants are the real wall.

• MM-GBSA rescoring: when physics beats the docking score

  MM-GBSA re-ranks docking poses with a physics-based endpoint. Here is what it does well, where it fails, and when it is worth the extra compute over a raw score.

• Why ΔΔ beats absolute docking scores

  Docking scores are bad at predicting binding affinity but good at ranking. Here is why the wild-type-minus-mutant difference is the number you should trust.

• Sonrotoclax: the first new BCL-2 inhibitor in a decade

  The FDA just approved sonrotoclax for relapsed mantle cell lymphoma, the first new BH3-mimetic since venetoclax. What it hits, and why molecular docking cares.

• EGFR C797S: the mutation behind osimertinib resistance

  C797S is the mutation that breaks osimertinib's covalent bond. Here's the mechanism, why allelic context decides treatment, and the fourth-generation inhibitors chasing it.

• The first PROTAC is approved: vepdegestrant in ESR1+ breast cancer

  The FDA approved vepdegestrant on May 1, 2026, the first PROTAC degrader ever cleared. What the VERITAC-2 data showed and why catalytic degradation matters.

• KRAS G12V: the common mutation still without a drug

  G12C has two approved drugs and G12D has clinical candidates, but KRAS G12V still has none. Here is the structural reason, and the pan-RAS programs trying to fix it.

• Plasma protein binding and the free drug hypothesis

  Only unbound drug reaches the target. Why chasing lower plasma protein binding usually wastes a med-chem program, and what fraction unbound really tells you.

• Boltz-2 and co-folding: when to use it instead of docking

  Co-folding models predict the protein-ligand complex from sequence and SMILES. Here is where Boltz-2 helps, where classical docking still wins, and how to combine them.

• KIT in GIST: the four-line resistance ladder

  Imatinib, sunitinib, regorafenib, ripretinib, avapritinib — how five drugs map onto KIT's primary and secondary mutations, and why each line eventually fails.

• PAINS: the compounds that fake their own activity

  What pan-assay interference compounds are, the chemistry behind their false hits, and why a PAINS flag is a prompt to confirm, not a verdict to discard.

• ROS1 fusion NSCLC: the TKI landscape in 2026

  A field guide to the four approved ROS1 inhibitors, why G2032R breaks most of them, and how the newest drugs were built to survive the solvent-front mutation.

• MET exon 14 skipping in NSCLC — capmatinib, tepotinib, and what skipping really means

  Why MET exon 14 skipping is an oncogenic event even without a kinase-domain mutation, how capmatinib and tepotinib exploit the resulting half-life increase, and where resistance shows up.

• PIK3CA H1047R and inavolisib — the mutant-selective degrader story

  Why PIK3CA H1047R activates PI3K alpha, how inavolisib goes beyond inhibition by degrading the mutant protein, and what the INAVO120 results mean for HR+/HER2- breast cancer.

• Blood-brain barrier and kinase inhibitors — what actually predicts CNS exposure

  Why most oral kinase inhibitors miss the brain, what physicochemical properties and efflux transporters predict CNS exposure, and how osimertinib, alectinib, and lorlatinib got it right.

• EGFR exon 20 insertions — the third resistance pocket

  Why classical EGFR TKIs fail on exon 20 insertions, and how amivantamab, sunvozertinib, and zipalertinib are building out a real treatment ladder for this 10% of EGFR-mutant NSCLC.

• BTK C481S and pirtobrutinib — the reversible bet that paid off

  How three generations of covalent BTK inhibitors hit the same C481S wall, why pirtobrutinib's non-covalent design escapes it, and what the docking actually shows.

• ProLIF interaction fingerprints — what your poses are actually doing

  Why an interaction fingerprint tells you more than a docking score, how ProLIF encodes the eight interaction types that matter, and where this fits in a screening workflow.

• KRAS G12D — drugging the bigger cousin in pancreatic cancer

  G12D is the most common KRAS mutation and the dominant driver in PDAC. With no cysteine to hook, the field had to invent two new playbooks. Here is where they stand.

• PROTACs and oral bioavailability — what bRo5 chemistry actually requires

  PROTACs blow past Lipinski on every count. Vepdegestrant still got oral. Here is the empirical bRo5 window that separates an oral degrader from an IV-only one.

• ALK G1202R: the solvent-front mutation that resets the ladder

  Why one ALK mutation knocks out crizotinib, ceritinib, alectinib and brigatinib at once, why lorlatinib still works, and what comes after it.

• Covalent docking: how it differs from the non-covalent kind

  Standard docking assumes reversible binding. For warhead-bearing drugs that form a real bond with the target, you need a different method. Here is how it works.

• DILI: the hepatotoxicity liability you can't fully design away

  Drug-induced liver injury is the leading cause of drug withdrawals. What causes it, why kinase inhibitors are prone, and how to flag it during molecular docking.

• HER2: the drug landscape across breast, gastric, and lung

  A field guide to HER2-amplified, HER2-low, and HER2-mutant cancer, the drugs that hit each, and where molecular docking fits the small-molecule TKIs.

• AlphaFold for docking: when the predicted structure is usable

  What the validation literature actually says about AlphaFold2 and AlphaFold3 structures as docking targets, and the pre-flight checks that catch the failure modes before they cost you a screen.

• FLT3 in AML: ITD, D835Y, and the type-I vs type-II pivot

  Why quizartinib works on FLT3-ITD but fails on D835Y, why gilteritinib doesn't, and what the F691L gatekeeper means for the next generation of FLT3 inhibitors.

• IDH1 R132H: one mutation, two diseases, two FDA approvals

  How a gain-of-function arginine-to-histidine swap in IDH1 produces an oncometabolite, drives both glioma and AML, and finally has clean drugs against it.

• PoseBusters: why your top-ranked pose can still be nonsense

  RMSD under 2 angstroms is not the same as a pose you can hand to a chemist. PoseBusters showed how often deep-learning docking violates chemistry, and what to check.

• BCR-ABL T315I: the gatekeeper that broke imatinib

  How a single threonine-to-isoleucine swap at position 315 walled off four BCR-ABL inhibitors, and how ponatinib and asciminib eventually got around it.

• CYP3A4: the metabolic bottleneck behind half your DDI risk

  Why CYP3A4 metabolizes most marketed oncology drugs, how strong inhibitors and inducers blow up exposure, and what to screen for early in lead optimization.

• BRAF V600E: same mutation, different drugs by tissue

  Why vemurafenib works in melanoma but fails in colorectal, how paradoxical RAF activation shapes the chemistry, and what BEACON changed about the standard of care.

• Vina, GNINA, and Glide: what each scoring function buys you

  An honest comparison of the three workhorse docking scoring functions in 2026 — what they actually optimize for, where they fail, and how to combine them.

• KRAS G12C — clinical landscape, 2026

  A field guide to the four approved KRAS G12C inhibitors, what's known about resistance pathways, and where the medicinal chemistry is going next.

• T790M, C797S, and the EGFR resistance staircase

  How EGFR resistance mutations evolved across three generations of inhibitors, what the docking actually shows, and why C797S is harder than T790M ever was.

• hERG: the silent killer of kinase programs

  Why a basic amine and a couple of aromatic rings are enough to wipe out a clinical candidate, and what the cardiac liability actually looks like in screening data.