Everything you need to run a binder design on Ranomics Tools, from first signup to sending a shortlist to the wet lab.
Seven steps from the protein you already have to a shortlist worth testing: work out which tool fits what is on your bench, open its page and read the form without an account, load its named pilot, sign in and submit, read the scores, then scale up or switch.
You fund a USD wallet and that balance is your cap. Design count is unlimited, no subscriptions. See how holds and final charges work.
Not sure which model fits your target. Browse all tools side by side with typical runtimes and what each one is built for.
What every tool is for, the inputs it needs, and how to read the scores it returns.
Upload your target structure, mark the residues you want gripped, and get back new mini-proteins of 60 to 150 residues built to grip them, already refolded and filtered. Sessions run up to four hours; results are emailed when they finish.
Upload your target, choose the format you want — mini-protein, nanobody, antibody or peptide — and get back candidates, each refolded from its sequence to check the fold holds and ranked on its predicted interface with your target. Handles sugars and modified residues on the target natively.
Choose your target, pick a small de novo binder or a paired heavy + light scFv (a single-chain antibody fragment), and get back designs ranked by a 0-to-1 interface confidence score in one model pass.
Upload your antigen and an antibody or nanobody, then redesign its binding loops, humanise its framework, raise its affinity, or just predict how the two dock — all aimed at the epitope you name.
Upload a protein or small-molecule target, name the chain and the residues you want gripped, and set how many designs to fund. The run fans out across GPUs and stops when your wallet does.
Upload your target structure, mark the residues you want gripped, and get back binders that each carry a real AlphaFold2 confidence score against that target. A pilot run takes roughly 30 to 60 min.
Upload your target structure, mark the patch you want gripped, and get back nanobody (single-domain antibody) candidates, each refolded and scored against the target.
Upload your target structure, mark the residues you want gripped, and get back brand-new binders, each carrying a real AlphaFold2 confidence score against your target. A pilot run takes roughly 15 to 30 min.
Upload a backbone — a structure with no sequence decided yet — and get back candidate sequences that should fold into it, each with a score and a native-recovery figure. About 30 s per run.
Paste a sequence — one chain or several — and get a predicted 3D structure back with per-residue and per-residue-pair confidence scores. About 5 to 10 min per run.
Paste a designed binder, upload the target it should hit, and get back the predicted complex plus a 0-to-1 confidence score for the contact between them. About 15 s per design in single-sequence mode.
Paste a sequence and get a predicted structure back in one to two minutes, with per-residue confidence. Trades a little accuracy for speed by skipping the search for related natural sequences.
Paste one protein sequence and get a predicted structure back in about 30 s, with per-residue confidence. One chain only, and no homolog search — so it works on designed sequences with no relatives.
Describe any mix of protein, DNA, RNA and small molecules in one spec and get the whole complex folded together, every atom modelled. The multi-molecule counterpart to the protein-only Boltz-2 tool.
These tools answer in seconds in the browser, with no GPU job to set up, so they have no separate guide — the page itself tells you what to paste in, and they open without an account: Epitope Scout, Binder Developability Scout
Billing, accounts, data and privacy, run times, what a design is, and what the scores mean.
Common problems and how to fix them, from PDB validation errors to a blank 3D viewer.
Still stuck or have a question the docs do not answer. Email the team and we will help.
