Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Analyze and engineer protein glycosylation. Scan sequences for N-glycosylation sequons (N-X-S/T), predict O-glycosylation hotspots, and access curated glycoengineering tools (NetOGlyc, GlycoShield, GlycoWorkbench). For glycoprotein engineering, therapeutic antibody optimization, and vaccine design.
.claude/skills/mkurman-glycoengineering/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 142% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 55% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 40% | 0% |
|----------|-------| | Enhance ADCC | Defucosylation at Fc Asn297 | Afucosylated IgG1 has ~50× better FcγRIIIa binding | | Reduce immunogenicity | Remove non-human glycans | Eliminate α-Gal, NGNA epitopes | | Improve PK half-life | Sialylation | Sialylated glycans extend half-life | | Reduce inflammation | Hypersialylation | IVIG anti-inflammatory mechanism | | Create glycan shield | Add N-glycosites to surface | Masks vulnerable epitopes (vaccine design) |
| Mutation | Effect | |----------|--------| | N297A/Q (IgG1) | Removes Fc glycosylation (aglycosyl) | | N297D (IgG1) | Removes Fc glycosylation | | S298A/E333A/K334A | Increases FcγRIIIa binding | | F243L (IgG1) | Increases defucosylation | | T299A | Removes Fc glycosylation |
| Symbol | Full Name | Type | |--------|-----------|------| | Glc | Glucose | Hexose | | GlcNAc | N-Acetylglucosamine | HexNAc | | Man | Mannose | Hexose | | Gal | Galactose | Hexose | | Fuc | Fucose | Deoxyhexose | | Neu5Ac | N-Acetylneuraminic acid (Sialic acid) | Sialic acid | | GalNAc | N-Acetylgalactosamine | HexNAc |
Typical complex biantennary N-glycan:
Neu5Ac-Gal-GlcNAc-Man\
Man-GlcNAc-GlcNAc-[Asn]
Neu5Ac-Gal-GlcNAc-Man/
(±Core Fuc at innermost GlcNAc)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 30,075 | 21,736 | -28% | 1 | 1 | 0% | 4,923 | 4,684 | -5% | 0 | 0 | — |
case-02 | fail→pass | 22,614 | 14,661 | -35% | 1 | 1 | 0% | 4,032 | 3,532 | -12% | 0 | 0 | — |
case-03 | pass→pass | 7,982 | 7,713 | -3% | 1 | 1 | 0% | 1,456 | 2,250 | +55% | 0 | 0 | — |
case-04 | pass→pass | 9,733 | 8,844 | -9% | 1 | 1 | 0% | 1,602 | 2,242 | +40% | 0 | 0 | — |
case-05 | pass→pass | 5,321 | 7,060 | +33% | 1 | 1 | 0% | 936 | 1,932 | +106% | 0 | 0 | — |
case-06 | pass→pass | 11,009 | 6,306 | -43% | 1 | 1 | 0% | 2,230 | 1,883 | -16% | 0 | 0 | — |
case-19 | pass→pass | 7,995 | 8,205 | +3% | 1 | 1 | 0% | 1,314 | 2,201 | +68% | 0 | 0 | — |
case-07 | pass→pass | 18,091 | 8,379 | -54% | 1 | 1 | 0% | 3,127 | 2,030 | -35% | 0 | 0 | — |
case-08 | fail→pass | 9,632 | 7,184 | -25% | 1 | 1 | 0% | 1,580 | 2,116 | +34% | 0 | 0 | — |
case-09 | pass→pass | 13,315 | 11,648 | -13% | 1 | 1 | 0% | 2,188 | 2,641 | +21% | 0 | 0 | — |
case-10 | pass→pass | 7,864 | 3,895 | -50% | 1 | 1 | 0% | 1,377 | 1,494 | +8% | 0 | 0 | — |
case-11 | pass→pass | 7,271 | 5,906 | -19% | 1 | 1 | 0% | 1,362 | 1,748 | +28% | 0 | 0 | — |
case-12 | pass→pass | 9,476 | 4,562 | -52% | 1 | 1 | 0% | 1,499 | 1,572 | +5% | 0 | 0 | — |
case-13 | fail→fail | 17,467 | 20,735 | +19% | 1 | 1 | 0% | 2,811 | 4,032 | +43% | 0 | 0 | — |
case-14 | pass→pass | 6,078 | 3,680 | -39% | 1 | 1 | 0% | 1,005 | 1,343 | +34% | 0 | 0 | — |
case-15 | pass→pass | 13,626 | 5,398 | -60% | 1 | 1 | 0% | 2,123 | 1,597 | -25% | 0 | 0 | — |
case-16 | pass→pass | 4,342 | 5,563 | +28% | 1 | 1 | 0% | 741 | 1,657 | +124% | 0 | 0 | — |
case-17 | fail→pass | 3,701 | 4,087 | +10% | 1 | 1 | 0% | 560 | 1,357 | +142% | 0 | 0 | — |
case-18 | pass→pass | 15,718 | 14,545 | -7% | 1 | 1 | 0% | 2,371 | 3,259 | +37% | 0 | 0 | — |
case-20 | pass→pass | 3,034 | 3,285 | +8% | 1 | 1 | 0% | 657 | 1,486 | +126% | 0 | 0 | — |
case-21 | pass→pass | 16,744 | 15,915 | -5% | 1 | 1 | 0% | 3,361 | 3,736 | +11% | 0 | 0 | — |
case-22 | pass→pass | 49,974 | 17,202 | -66% | 1 | 1 | 0% | 4,111 | 3,920 | -5% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +14 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.