Research photograph
Research & development
PepSilico and peptide-science research activities.
PepSilico webScavenPred and ACE-iPred are available as standalone research prediction tools. SIPAD remains in development.
PEPTIDE COMPUTATION · DISCOVERY · INTELLIGENCE
Computational Intelligence
for Peptide Science
PepSilico is a scientific environment for peptide activity prediction, screening, sequence intelligence and multifunctional peptide discovery.
Initial activity systemsACE inhibitionAntimicrobial activity
SEQUENCE · INTELLIGENCE · DISCOVERYAYHTIK
Sequence→Evidence→Activity
PredictScreenDiscoverTrain
PUBLIC PREDICTION TOOLS
ScavenPred & ACE-iPred
Standalone published-model prediction services. Enter a peptide sequence or paste multiple sequences for batch screening.
Use the 20 standard amino-acid one-letter codes. One sequence per line.
PEPTIDE ACTIVITY DISCOVERY
SIPAD
Sequence Intelligence for Peptide Activity Discovery
One environment for individual prediction, library screening and discovery of peptides with multiple predicted functions.
SINGLE / MULTI-ACTIVITY
GUI PREVIEWActivity Prediction
Enter a peptide and select the biological activities to interrogate independently.
Standard one-letter amino-acid notation
PEPTIDE LIBRARY
Batch Screening
Screen multiple peptide sequences across one or more selected activities.
One sequence per line
OR
FUNCTION INTERSECTION
Multifunctional Discovery
Define the combination of functions of interest and filter a peptide library accordingly.
ACE+Predicted positive
∩AMP+Predicted positive
=MultifunctionalCandidate set
USER-DEFINED ENDPOINT
Trainable Model
Reserved for the existing trainable architecture and labelled peptide datasets.
1Provide datasetSequences + labels
2Build knowledgeTrainable architecture
3ValidateInspect performance
4PredictInterrogate new sequences
Scientific integration pendingThis GUI does not recreate or alter the trainable engine. It will be connected later from the verified implementation.
SCREENING OUTPUT
Results & Downloads
Designed for complete prediction matrices and multifunctional filtering.
| Sequence | ACE | AMP | Multifunctional | Actions |
|---|---|---|---|---|
| PEPTIDE_01 | Awaiting engine | Awaiting engine | — | |
| PEPTIDE_02 | Awaiting engine | Awaiting engine | — |
POSITIVE REASONING CARD
Sequence
Verified engine reasoning will appear here.
NEGATIVE REASONING CARD
Sequence
Verified engine reasoning will appear here.
MEDIA · RESEARCH IN ACTION
Research in action
Selected moments from PepSilico research, development and scientific engagement.
PUBLICATIONS
Research associated with PepSilico
This section will provide the formal scientific record for PepSilico, SIPAD and associated activity systems. Only published or otherwise publicly citable research will be listed with verified bibliographic information.
Publication record in preparationNo unpublished manuscript is presented here as a published work. Verified titles, journals, DOI links and citation formats will be added when they become publicly citable.
Sequence-intelligence research
Research describing the underlying position-aware, evidence-reasoning approach and its evaluation.
Bibliographic details pending public releaseActivity-specific studies
Peer-reviewed studies and associated resources applying PepSilico/SIPAD methods to peptide activity problems.
Bibliographic details pending public releaseHow to cite PepSilico
A recommended software/platform citation and version identifier will be supplied before the public scientific release.
Citation record not yet assignedUPDATES
Platform development
Release notes will document public changes to PepSilico, SIPAD and individual activity systems so users can distinguish interface updates from scientific-engine changes.
CURRENT · DEVELOPMENT
Interface development
PepSilico is currently being prepared as a scientific web environment. The interface includes activity prediction workspaces, batch and multifunctional screening, research resources, documentation and evidence-contribution workflows.
Scientific engines are not yet publicly connected.PLANNED · VERIFIED INTEGRATION
Activity-system integration
Validated activity engines will be connected only after interface development, followed by regression checks against their frozen reference outputs.
Release date not yet assigned.PLANNED · PUBLIC RELEASE
Versioned scientific release
Public releases will identify relevant interface and scientific-component versions, supported activity endpoints and associated documentation.
Release date not yet assigned.RESOURCES
Scientific resources
Access documentation, approved datasets, tutorials and downloadable materials associated with PepSilico and SIPAD. Resources will be released progressively as individual scientific components are validated for public use.
▤
DOCUMENTATIONSIPAD User Guide
Guidance for prediction, batch screening, multifunctional discovery, reasoning outputs and experimental feedback.
Read help guide →◫
DATADatasets
Approved training, benchmark or supplementary datasets can be listed here with endpoint definitions, provenance and version information.
View datasets →↓
DOWNLOADSResearch Downloads
Publication supplements, approved prediction outputs, figures and supporting research material.
Release controlled▷
TUTORIALSWorked Examples
Step-by-step examples showing how to submit sequences, interpret outputs and screen for multiple peptide functions.
View workflow →Versioned scientific resourcesPublic datasets and model-associated resources should identify the relevant activity endpoint and version so that results remain reproducible.
SIPAD HELP & DOCUMENTATION
Using SIPAD
This guide describes the interface behaviour. Prediction controls remain inactive until the verified scientific engines are integrated.
Activity Prediction
Enter a peptide using standard one-letter amino-acid notation, then select one or more available activity systems. Each selected activity is interrogated independently; selecting multiple activities does not merge their knowledge bases.
Batch Screening
Paste multiple peptide sequences or, after backend integration, upload a supported file. SIPAD is designed to return a prediction matrix containing the outcome from each selected activity system for every submitted peptide.
Multifunctional Discovery
Multifunctional discovery filters the independent outputs of selected activity systems. For an ALL-functions search, a peptide is retained only when it satisfies the required prediction outcome for every selected activity.
ACE+ ∩ AMP+ → candidate multifunctional peptides
Understanding Results
Activity systems may return Positive, Negative or Unresolved according to the evidence available to the corresponding engine. Results from different activity systems should be interpreted separately before applying any multifunctional filter.
PositiveNegativeUnresolved
What does “Unresolved” mean?
Unresolved is an intentional outcome, not an automatically assigned negative class. It indicates that the available admissible evidence did not establish sufficient decision authority for that query under the relevant activity system.
Reasoning Cards
Where supported by the connected engine, the interface will present the actual reasoning information returned for a prediction. Positive outputs use the blue-card convention and Negative outputs use the red-card convention. The website will not invent reasoning content that is absent from the scientific engine.
Report an Experimental Outcome
Experimentally tested predictions can be returned through the Contribute area. Users may report either confirming or discrepant outcomes and provide endpoint values, units, assay context and supporting references. Submissions enter scientific review; they do not automatically modify an activity model or knowledge base.
Go to Contribute →1Enter sequence(s)
→2Select activity
→3Run prediction
→4Review reasoning
→5Download or contribute
COMMUNITY EVIDENCE
Contribute peptide evidence
Help strengthen the scientific evidence base by returning experimentally tested predictions or submitting newly identified peptides. Submitted evidence is intended for review and curation; submission does not automatically alter a prediction model or knowledge base.
↩
VALIDATION FEEDBACK
Open form →
Report Prediction Outcome
Return an experimentally tested SIPAD peptide, including outcomes that disagree with or confirm the prediction.
＋
NEW EVIDENCE
Open form →
Submit Newly Identified Peptide
Submit a newly identified, experimentally characterised peptide for scientific review and possible future curation.
RETURN EXPERIMENTAL EVIDENCE
CURATION REQUIREDReport Prediction Outcome
Use this form after experimentally testing a peptide that was screened with SIPAD.
NEWLY IDENTIFIED PEPTIDE
CURATION REQUIREDSubmit New Peptide Evidence
Provide experimental evidence for a peptide that may not yet be represented in PepSilico.
1Submission
→2Scientific review
→
3Evidence audit
→4Curated eligibility
RESEARCH
Position-aware intelligence for peptide science
PepSilico provides a home for research into peptide sequence reasoning, activity prediction and multifunctional discovery. The research programme treats peptide activity as a sequence problem in which residue identity and positional organisation are both preserved during evidence interrogation.
Residue identity+Positional organisation=Sequence information
Preserving sequence organisation
Closely related peptides can contain similar residues while differing in positional organisation. PepSilico research therefore emphasises prediction frameworks in which sequence position remains explicit rather than treating compositional similarity as functional equivalence.
Evidence before decision
The underlying research explores interpretable reasoning in which admissible positional evidence is progressively resolved for an individual query. The aim is not merely to return a class, but to retain an inspectable relationship between the available evidence and the resulting decision.
Independent biological endpoints
Activity-specific systems are maintained independently. Initial PepSilico work includes ACE inhibitory peptide prediction and antimicrobial activity against Staphylococcus aureus. Multifunctional discovery operates across outputs of independent activity systems rather than combining their evidence bases.
Prediction, validation and curation
PepSilico is being designed to support the return of experimentally tested predictions and submission of newly identified peptides. New evidence is subject to scientific review and audit before it can become eligible for future curated use.
Research principle
Similarity can be informative without implying functional equivalence. PepSilico is being developed to preserve peptide identity while allowing evidence relationships to be interrogated in a controlled and traceable way.
PRIVACY · DATA USE · TERMS
Responsible use of PepSilico
These pre-release notices establish the intended handling principles for the platform. They are not a substitute for the final institutional/legal review required before public deployment.
Research-use notice
PepSilico predictions are computational research outputs. They should not be treated as clinical, diagnostic or therapeutic advice, or as a substitute for experimental validation.
Prediction uncertainty
A prediction represents the output of the selected activity-specific system under its defined endpoint and version. An Unresolved result indicates insufficient decision authority under that system; it should not be reinterpreted as a Negative prediction.
Submitted peptide data
Experimental outcomes and newly identified peptide records submitted through Contribute are intended to enter a review workflow. Submission alone does not make evidence part of a model or knowledge base.
Evidence curation
Submitted scientific evidence may require provenance, endpoint, assay and supporting-reference checks before it is considered eligible for curated research use.
Personal information
The public submission workflow should collect only information necessary for scientific communication and provenance. Final retention periods, access controls, contact handling and deletion procedures will be specified before submissions are enabled.
Intellectual property
Before public submission is enabled, contributor terms should state how submitted sequences, experimental results and supporting materials may be reviewed, stored, cited or reused. Users should not submit confidential or restricted material unless the final terms explicitly permit it.
Before public launchFinal Privacy Policy, Terms of Use, contributor/data-use terms, institutional contact details, governing jurisdiction and any required ethics or data-governance statements must be reviewed and published.
ABOUT PEPSILICO
A scientific environment for computational peptide discovery
PepSilico is a developing research platform for computational peptide science. It brings peptide activity prediction, library screening, multifunctional discovery, scientific resources and evidence contribution into a single environment.
Computational Intelligence for Peptide Science
The platform is intended to connect computational prediction with experimentally grounded peptide research while keeping individual biological activity systems scientifically distinct.
PredictInterrogate peptide sequences using activity-specific systems.
ScreenEvaluate peptide libraries efficiently across selected functions.
DiscoverIdentify candidate multifunctional peptides from independent predictions.
ContributeReturn experimental outcomes and newly characterised peptide evidence for review.
PLATFORMPepSilicoScientific ecosystem
→DISCOVERY ENVIRONMENTSIPADSequence Intelligence for Peptide Activity Discovery
→ACTIVITY SYSTEMSACE · AMP · future modulesIndependent endpoint-specific prediction
DEVELOPMENT
Research-led development
PepSilico is being developed from an ongoing peptide sequence-intelligence research programme. Formal institutional affiliation, contact information, citation instructions and publication links will be added before public release.
Pre-release statusThe current website is a GUI development build. Scientific engines and user-submission services are not yet publicly connected.