Refining Single Leads. Advancing Discovery.
Leverage data-driven insights to assess potential, optimize leads, and advance promising candidates.
Conceptual workflow. The dashed stages happen before this service: the lead you already hold is produced elsewhere, not here.
Evaluates. The one antibody lead you already hold, with its sequence, chain information and any metadata or objectives you supply.
Why it matters. The lead anchors the work. Everything that follows is measured against it.
May produce. A defined starting point and optimization objective.
What this service does
- 01Start with a single leadBegin with an existing candidate and establish a clear reference for optimization.
- 02Explore variantsPropose sequence variants of the parent and order them by sequence-based descriptors, for testing by your team.
- 03Prioritized candidatesCompare computational scores and identify promising candidates for further evaluation.
Follow one lead through the analysis
Select a stage. The example below uses an illustrative sequence, and stage 3 happens only when a structure is supplied.
Start with a Single Lead
Input and optimization objective
Analyze the given single lead sequence. Establish the starting sequence, define the desired improvements, and set a reference for evaluating potential modifications.
- Considers
- Original sequence, available metadata, and optimization goals.
- Output
- AI-generated and optimized library from a single lead.
Illustrative example sequence.
Which service is this? Single Lead Optimization works on variants of one lead. To compare several leads or a dataset, see Lead Optimization. To start with no lead, see De Novo.
From Inputs to Insights
- Lead sequence (FASTA)
- Chain information, such as VH and VL
- What you want to improve
- Structure or experimental data, if you have it
- Scores are used to rank or organize variants.
- Candidate variants are computational proposals, not experimentally confirmed improvements.
- Structural analysis depends on the availability and scope of structural data.
- Heuristic scores do not establish binding, stability, expression or efficacy.
Read scientific limitations
A score orders variants under one heuristic. It is built from substitution patterns and residue properties, so it knows nothing about your target, your assay or your expression system.
A variant that ranks highly may still lose binding or fail to express. A variant that ranks low may work well. Only experiments can tell.
Structural context depends on a structure being supplied and on how well it represents your lead. Where none is supplied, the analysis is sequence-level only.
Evaluate your lead candidate.
Tell us about your sequence, any structure you hold, and what you want to improve. We will discuss which computational workflow fits.