Feature tutorials
What each platform feature does, when to use it and when not to. Open any card for the full walkthrough.
The platform now follows a workspace-first R&D loop: ingest real wet-lab diversity, triage it into a shortlist, score structures and affinity, review project-specific risks, then hand a candidate packet back to the bench.
- 1.Intake wet-lab diversity — NGS Pool ingests phage-display / functional-sort sequence pools and keeps clone counts, tiers, and target metadata attached.
- 2.Triage the pool — Semi-de-novo Triage reduces ~10^6 sequences to a ranked ~10^2 shortlist using cheap sequence gates before expensive structure scoring.
- 3.Triage generated poses — Interface predicts antibody-antigen complex poses and extracts geometry/confidence readouts, PDBs, and CSV artifacts; these do not establish binding.
- 4.Rank affinity and readiness — Affinity Engine shows model readiness, target coverage, candidate scoring, and maturation scans where calibrated data exists.
- 5.Review project-specific signals — G-protein Design-around and CAR Interface Beta provide envelope-risk and dual-target CAR pose-triage signals without changing the standard Interface workspace.
- 6.Hand off a candidate packet — Runs & Candidates, Interface Jobs, Compare, and exports collect provenance, artifacts, metrics, and caveats for wet-lab discussion.
Important caveat: platform scores are evidence for prioritization, not final biological truth. Wet-lab KD, specificity, expression, function, and safety checks still decide.
Workspace
Pick a research task, read the whole plan before anything runs, approve a compute ceiling, and stop at the decision point the plan puts in front of the expensive half.
Use when: You want a complete design-and-rank chain rather than a single tool run, and you want to see the cost and the tool list before spending anything.
Every tool the platform can call, filtered by modality and function, each labelled by where it runs and how far it has been verified.
Use when: You want to know what is available for a modality, or to check what a plan step will actually call before you approve it.
One target per card: its library, its measured cohort, its predictions — and an explicit flag when there is no measured cohort to calibrate against.
Use when: Before trusting any score for a target, and when deciding which target is ready for a scoring-driven campaign.
Sequence first, epitope second. Start from a known target, recover antibody/antigen sequences and SEQ IDs, then evaluate patent/public evidence and optional epitope disclosure.
Use when: You know the target and need to find therapeutic antibody evidence, disclosed VH/VL/CDR/SEQ IDs, antigen sequences, clones, patent families, explicit or implicit epitope protection, and public evidence that can guide design-around or FTO evidence-packet preparation.
Ingest wet-lab phage-display and functional-sort sequence pools, dedupe VH clones, and expose per-tier read counts for triage.
Use when: A lab has delivered FASTA/CSV/TSV sequence reads or a manifest with sample tiers and you need the pool available for semi-de-novo triage.
Funnel a real wet-lab NGS pool from about 10^6 sequences to a ranked shortlist of about 10^2 candidates.
Use when: You have an enriched sequence pool and need a bench-usable shortlist for expression or validation.
Predict antibody-antigen complexes from sequences and inspect ipTM, contact residues, BSA, pDockQ, PDBs, and CSV exports.
Use when: You have one or many antibody sequences and need a structure-level readout against a target antigen.
The operations table for every predicted complex: filter, reopen, select completed jobs, export CSV/PDB/contact artifacts, and compare epitope fingerprints.
Use when: You need to find an old job, export many completed jobs, delete failed clutter, or compare contact-residue fingerprints.
One workspace for execution history and candidate library review.
Use when: You need to check progress, reopen outputs, inspect candidates, or hand off a ranked set.
Per-target affinity scoring, maturation scans, calibration status, and wet-lab readiness in one workspace.
Use when: You need target-calibrated ranking, a maturation scan, or a reality check on whether affinity predictions are ready for a given target.
Upload a protein complex, assign receptor and ligand chains, run MD + MM/PBSA on Modal, and inspect energy results and trajectory artifacts in one workspace.
Use when: You have a standard-residue protein–protein PDB/ENT complex and need a technically reproducible relative binding-energy estimate from a matched MD protocol.
Technical triage for alternative pseudotyping glycoproteins against VSV-G / Cocal-G risk anchors.
Use when: A lentiviral envelope program needs alternatives or a 354-position risk discussion must be made reproducible.
A beta 2 x 2 generated-pose triage for isolated arms in dual-target CAR concepts.
Use when: You have two binder arms and two antigen ECD sequences and need early structural triage before wet-lab or construct work.
Specialized
Structure-based antibody/nanobody design runner with framework presets, hotspot input, gates, job history, and artifact downloads.
Use when: You have a target PDB, hotspot residues, and want RFdiffusion-Ab/ProteinMPNN/RF2-Ab style design artifacts.
A WIP cassette-planning surface for binder, hinge, TM, costim, and CD3z choices.
Use when: You need to discuss construct architecture and component choices after a binder has been shortlisted.
Interactive residue-patch selection and visualization for target structures.
Use when: You need to define or verify the antigen patch before RFantibody, older pipeline flows, or manual review.
Legacy orchestrated design launcher. It remains available for specialized runs, but the current day-to-day platform path is Workspace-first.
Use when: A project specifically needs this format, constraint, or legacy flow and the responsible owner has confirmed the backend path for the run.
Specialized escape-mutation review surface for checking whether a binder may retain signal under antigen substitutions.
Use when: A project specifically needs this format, constraint, or legacy flow and the responsible owner has confirmed the backend path for the run.
Power-user constraint-design UI for non-standard generative setups. Use only when standard workspace modules are too narrow.
Use when: A project specifically needs this format, constraint, or legacy flow and the responsible owner has confirmed the backend path for the run.
VHH-specific design surface for single-domain antibody work. Backend handoff should be checked before treating outputs as production runs.
Use when: A project specifically needs this format, constraint, or legacy flow and the responsible owner has confirmed the backend path for the run.
Tandem/dual-VHH CAR concept surface. Use for architecture planning, then inspect isolated-pair pose signals in CAR Interface Assessment and verify binding experimentally.
Use when: A project specifically needs this format, constraint, or legacy flow and the responsible owner has confirmed the backend path for the run.
Small protein binder concept surface for compact formats. Route any serious hit through Interface and Affinity before review.
Use when: A project specifically needs this format, constraint, or legacy flow and the responsible owner has confirmed the backend path for the run.
Specialized lentiviral transgene optimization surface for packaging-titer and expression-safety discussions.
Use when: A project specifically needs this format, constraint, or legacy flow and the responsible owner has confirmed the backend path for the run.
Utilities
A focused sequence-to-structure UI over per-model run endpoints.
Use when: You have an existing antibody sequence and need a quick structure for inspection or downstream review; choose ImmuneBuilder.
Standalone Mol* viewer for inspecting PDB files and extracting per-chain sequences.
Use when: You need to quickly view a PDB, check chain layout, or extract sequence before using another module.
Side-by-side or overlay comparison for completed structures and Interface jobs.
Use when: You need head-to-head visual review between candidate structures, competitor complexes, or Interface jobs.
Registry of integrated models and single-tool run surfaces.
Use when: Looking for a specific backend, checking model capabilities, or running a one-off model task.