What can I use it for?
Use the two tools to explore research hypotheses; validate predictions experimentally and cite the paper.
Why is my job waiting?
One CPU worker processes accepted jobs in order. Queue length and sequence encoding affect waiting time.
How much can I submit?
Up to 100 pairs, 10,000 residues per sequence, and 5 MiB per request. Retrieval accepts one query and Top K of 5, 10, or 20.
Are long sequences fully processed?
No. ESM uses at most 512 tokens including special tokens; AAindex uses the first 100 residues.
Is similarity an interaction probability?
No. Cosine similarity ranks natural candidates. Run a separate classification for a pair probability.
Does retrieval generate proteins?
No. Every returned sequence comes from the fixed natural reference library.
Do I need an account?
No. Access uses the private job link; anyone with its full token can access the job.
How long are results kept?
Seven days from submission. Access stops at expiry and server data is cleaned within one hour.
Can I recover a lost link?
There is no account or public job list for recovery. Keep the full link and download results before expiry.
How do I cite the tool?
Use the full DeepHVI citation on the citation page, with DOI and BibTeX.