FAQ

It depends on the question. Target and indication questions can be answered entirely from public databases such as TCGA, ICGC, GEO and CCLE. If the question involves your own samples, we work with scRNA-seq, bulk RNA-seq, WES, ctDNA and clinical sample data. We'll tell you exactly what's needed during the free 30-minute consultation.

Yes. The Target–Indication Evidence Pack, for example, is built entirely from public databases and needs none of your internal data to start. Many target assessment and indication questions can be read from public data before you commit lab resources.

Your team certainly could. The difference is in three places. Consistency: across several targets or projects, results are only comparable if they use the same format and the same methods. Traceability: every deliverable comes with reproducible scripts, and every number can be followed back to its source, so it holds up when the other side checks it. Interpretation: results have to be mapped back to mechanism and clinical context before they can support a decision.

Alongside them, not instead of them. We usually take on the part your team doesn't have time or specific experience for: single-cell analysis, external validation of a predictive model, or a burst of work before a milestone. Code ships with every report, so your team can review the work and carry it forward.

Four commitments. Your data stays on your servers: we work through remote access. Confidential from day one: an NDA is signed before kickoff and binds the whole team. Returned or destroyed: all copies at project close. Compliance first: we follow human genetic resources and data security regulations, and client data never goes into public models. See Data security.

When your data is involved, we connect to your environment remotely and run the analysis on your servers. The data doesn't move. Public-data projects don't require access to your systems at all. Discussions and results walkthroughs happen over video calls.

It depends on the service. Deliverables include reports, publication-grade figures, sample-level results tables, QC reports, statistical analysis plans, model code with validation results, nomograms or risk stratification tools, and reproducible scripts. Figures and conclusions are prepared for the meeting room and the data room.

It depends on scope and the state of the data. We give you a timeline at the proposal stage. A Target–Indication Evidence Pack takes about 4–6 weeks, depending on scope.

After the free consultation, we send a proposal and quote based on the question and the data. Pilots are quoted at a fixed scope. Pricing for larger target sets or ongoing work is discussed separately.

Yes. If the report is going to an international partner or an English-language journal, let us know at the start.

Our bases are in China, Europe and Taiwan. We agree meeting times with you at kickoff.

No. AI speeds up the work: literature and clinical trial database scans, code generation and report drafting. A computational biologist checks every number, and every conclusion traces back to its source. We don't do black-box predictions, we don't promise that "AI will find your target," and client data never goes into public models.

Yes. Manuscript support is part of what we do, and our team members have co-authored with partners before, for example the PLT012 study in Cancer Discovery 2025, with author contributions in data curation, formal analysis and methodology. Authorship follows actual contribution and is agreed with you in advance.

No. Analysis of public data answers disease-side questions: where the target is specifically expressed, where the normal-tissue risks are, how it relates to immune infiltration and prognosis, and which cell types express it. It can't establish properties of the molecule itself, such as epitope or affinity. Each report states the scope of every conclusion.