Bennett 2025 Supplementary Information, including Supplementary Methods Tables 5 and 6

What this source contributed

Records in this release that cite this source. A record can cite several sources.

• 1 configuration
• 4 protocols
• 7 datasets
• 30 evaluations
• 36 results
• 18 evidence claims
• 1 use case

Results

Values as the source printed them. Compare results only under compatible protocols, datasets and metrics.

Claims

Statements this source supports, each with where in the source it was checked.

• CDR-H3 mean sequence similarity of H1 stem binders (96-design row): Cdr h3 mean sequence similarity of binders 33.4% (Supplementary Methods Table 6, row 'H1 stem' (Designs screened = 96), column 'CDR-H3 mean sequence similarity of binders (%)')
• CDR-H3 mean sequence similarity of H1 stem binders (9000-design row): Cdr h3 mean sequence similarity of binders (33.4%) (Supplementary Methods Table 6, row 'H1 stem' (Designs screened = 9000), column 'CDR-H3 mean sequence similarity of binders (%)')
• CDR-H3 mean sequence similarity of IL7R binders (96-design row): Cdr h3 mean sequence similarity of binders N/A (Supplementary Methods Table 6, row 'IL7R' (Designs screened = 96), column 'CDR-H3 mean sequence similarity of binders (%)')
• CDR-H3 mean sequence similarity of RSV site 3 binders (9000-design row): Cdr h3 mean sequence similarity of binders N/A (Supplementary Methods Table 6, row 'RSV site 3' (Designs screened = 9000), column 'CDR-H3 mean sequence similarity of binders (%)')
• CDR-H3 mean sequence similarity of SARS-CoV-2 RBD binders (9000-design row): Cdr h3 mean sequence similarity of binders 20.4% (Supplementary Methods Table 6, row 'SARS-CoV-2 RBD' (Designs screened = 9000), column 'CDR-H3 mean sequence similarity of binders (%)')
• CDR-H3 mean sequence similarity of TcdB binders (96-design row): Cdr h3 mean sequence similarity of binders 22.2% (Supplementary Methods Table 6, row 'TcdB' (Designs screened = 96), column 'CDR-H3 mean sequence similarity of binders (%)')
• OrthoRep affinity-maturation fold improvement for H1 stem (96-design row): Orthorep fold improvement in affinity N/A (Supplementary Methods Table 6, row 'H1 stem' (Designs screened = 96), column 'OrthoRep fold improvement in affinity')
• OrthoRep affinity-maturation fold improvement for H1 stem (9000-design row): Orthorep fold improvement in affinity ~1000x (Supplementary Methods Table 6, row 'H1 stem' (Designs screened = 9000), column 'OrthoRep fold improvement in affinity')
• OrthoRep affinity-maturation fold improvement for IL7R (96-design row): Orthorep fold improvement in affinity N/A (Supplementary Methods Table 6, row 'IL7R' (Designs screened = 96), column 'OrthoRep fold improvement in affinity')
• OrthoRep affinity-maturation fold improvement for RSV site 3 (9000-design row): Orthorep fold improvement in affinity N/A (Supplementary Methods Table 6, row 'RSV site 3' (Designs screened = 9000), column 'OrthoRep fold improvement in affinity')
• OrthoRep affinity-maturation fold improvement for SARS-CoV-2 RBD (9000-design row): Orthorep fold improvement in affinity ~1000x (Supplementary Methods Table 6, row 'SARS-CoV-2 RBD' (Designs screened = 9000), column 'OrthoRep fold improvement in affinity')
• OrthoRep affinity-maturation fold improvement for TcdB (96-design row): Orthorep fold improvement in affinity ~100x (Supplementary Methods Table 6, row 'TcdB' (Designs screened = 96), column 'OrthoRep fold improvement in affinity')
• Relevance of uc-binder-20261022-protocol-bennett-insilico-af3-success to "Which computational design workflow produces binders that work when they are made and tested?": Assessed by Bennett VHH computational success rate under an AlphaFold3 filter (uc-binder-20261022-source-bennett-si: Supplementary Methods Table 6, column 'Success rate (%) (AF3)')
• Relevance of uc-binder-20261022-protocol-bennett-insilico-rf2-success to "Which computational design workflow produces binders that work when they are made and tested?": Assessed by Bennett VHH computational success rate under the fine-tuned RoseTTAFold2 filter (uc-binder-20261022-source-bennett-si: Supplementary Methods Table 6, columns 'Success rate (%) (RF2, hotspot proportion = 1 / 0.1 / 0.0)')
• Relevance of uc-binder-20261022-protocol-bennett-vhh-ecoli-spr-96 to "Which computational design workflow produces binders that work when they are made and tested?": Assessed by Bennett VHH design, E. coli expression and single-concentration SPR screen of 96 designs (uc-binder-20261022-source-bennett-si: Supplementary Methods Table 6, rows with 'Initial screening method' = SPR; uc-binder-20261022-source-bennett: Supplementary Methods Table 6, rows with 'Initial screening method' = SPR)
• Relevance of uc-binder-20261022-protocol-bennett-vhh-yeast-display-9000 to "Which computational design workflow produces binders that work when they are made and tested?": Assessed by Bennett VHH design, yeast surface display screen of 9,000 designs with SPR confirmation (uc-binder-20261022-source-bennett-si: Supplementary Methods Table 6, rows with 'Initial screening method' = YSD; uc-binder-20261022-source-bennett: Supplementary Methods Table 6, rows with 'Initial screening method' = YSD)
• Summary of the de novo binder design evidence: Summary Four studies, pinned here as seven artifacts, carry measured outcomes for de novo binder design, and none of them answers which workflow to use, because no paper runs two design methods and a conventional comparator on the same targets with one assay and one hit definition. The literature pass behind this evidence found nine candidate sources with measured outcomes; the five that were not retrieved are recorded as gaps rather than as evidence of what they would show. What the measured evidence shows is that the target matters more than the method. AlphaProteo's yeast-display success rate across eight targets runs from 0% over 54 designs on TNF-alpha to 88% over 94 designs on BHRF1, at a binding-signal threshold of 0.2 (1.3 on IL-17A). BindCraft reports an average experimental success rate of 46.3% over twelve targets with a range of 10% to 100%, built from per-target counts as small as 2 binders from 7 designs on Bet v1 and as large as 6 from 6 on the SpCas9 REC1 domain. The same family of methods applied to antibody variable domains is two orders of magnitude worse: fine-tuned RFdiffusion gives 2% over 96 designs on TcdB, 0% over 96 on IL-7 receptor alpha, and 0.01% to 0.2% over 9,000 designs per target by yeast display. On influenza H1 stem the same design campaign gives 1% over 96 designs by SPR and 0.2% over 9,000 by display, so the screening method moves the rate as much as the target does. Nearly all of this is author-reported by the group that built the method, which also chose the targets, ran the assay and chose which designs to test. The exception is the Adaptyv EGFR competition, where a third-party laboratory expressed and measured designs from many entrants on one target and one assay: 2.5% of 201 round-1 designs bound (printed elsewhere in the same paper as 3%), and 53 of 400 round-2 designs bound, printed as 13%. That source cannot be turned into a per-method rate: design methods are the entrants' own declarations, the organisers hand-picked 101 of 201 round-1 and 300 of 400 round-2 designs for diversity, and the three binding thresholds behind the figure counts are never defined. Affinities reached are better than the hit rates suggest. AlphaProteo's best measured binder was 0.082 nM on IL-7 receptor alpha by HTRF; BindCraft's best ranged from 5 nM on CbAgo to 5.7 micromolar on CrSAS-6; the best de novo design in the Adaptyv competition reached 82 nM and the winning optimised cetuximab fragment 1.21 nM, against 9.94 nM for the cetuximab positive control measured on the same instrument. Three things stop these numbers from ranking methods. The hit definitions are different endpoints, not different thresholds on one endpoint. The denominators are small and chosen, and the designs tested are the top picks of each method's own filter, so design and selection are coupled: changing only AlphaProteo's filter moved its PD-L1 rate from 9.6% over 94 designs to 16.1% over 31, and the Bennett authors attribute their low antibody rates partly to a filter they describe as weak. And the spread within one method across targets is wider than any plausible difference between methods. No conventional comparator is run anywhere in this evidence, so there is no measurement of what a designed binder is being compared against. (AlphaProteo Table 1, Table S2 and Sections S1.2.1, S2 and S4.1; BindCraft Introduction and Results prose; Bennett Supplementary Methods Table 6; Adaptyv Sections 2.3, 2.4, 3.1, 3.2, 4.1, 4.2 and 4.3)
• Two Bennett VHH design targets have no outcome row: Targets without printed outcome Supplementary Methods Table 5 rows: HIV Env (2NY7), SARS-CoV-2 RBD (6M0J), RSV-F Site I (7LVW), RSV-F Site III (7LUC), Influenza Hemagglutinin (5VLI), TcdB (6C0B), IL-7 Receptor alpha (3DI3). Section 6.8 states 'We designed VHHs against 6 targets in total', which does not match the seven rows. Table 6's caption states 'For 5 targets for which de novo VHH binders were identified, details are provided', and Table 6 has rows only for TcdB, IL7R, H1 stem (twice), SARS-CoV-2 RBD and RSV site 3. No outcome is printed for RSV-F site I or HIV Env, and none is inferred here. (Supplementary Methods Tables 5 and 6; Section 6.8)

Models, benchmarks and other records

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.