in-silico-perturbation-oracle

Virtual gene knockout simulation using foundation models to predict transcriptional changes

Install with Tessl CLI

npx tessl i github:aipoch/medical-research-skills --skill in-silico-perturbation-oracle

What are skills?

Review — 27%

Does it follow best practices?

If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:

npx tessl skill review --optimize ./path/to/skill

Learn more

Validation — 10 / 11 Passed

Validation for skill structure

SKILL.md

Review

Evals

Evaluation results

100%

58%

Oncology Target Prioritization Report

Target scoring output format and weights

Criteria

Without context

With context

target_gene column

100%

efficacy_score column

100%

safety_score column

100%

druggability_score column

100%

novelty_score column

100%

overall_score column

100%

recommendation column

100%

Efficacy weight 0.35

100%

Safety weight 0.25

100%

Druggability weight 0.25

100%

Novelty weight 0.15

100%

Standard cell type name

100%

Valid perturbation_type

100%

Without context: $0.2357 · 1m 19s · 14 turns · 18 in / 4,526 out tokens

With context: $0.9241 · 2m 51s · 35 turns · 1,699 in / 7,426 out tokens

87%

43%

Transcriptomics Pipeline for Cardiac Gene Perturbation Study

Differential expression and pathway enrichment output format

Criteria

Without context

With context

gene_symbol column

100%

log2_fold_change column

100%

p_value column

100%

adjusted_p_value column

100%

perturbed_gene column

100%

cell_type column

100%

pval threshold 0.05

100%

logfc threshold 1.0

100%

Wilcoxon method

fdr_bh correction

70%

Pathway JSON database key

100%

Pathway entry fields

30%

100%

KEGG or GO_BP included

100%

Without context: $0.4224 · 2m 8s · 18 turns · 23 in / 7,571 out tokens

With context: $0.8739 · 3m 11s · 31 turns · 40 in / 10,718 out tokens

55%

37%

Synergy Screening for Combination Immunotherapy Targets

Combinatorial knockout and QC best practices

Criteria

Without context

With context

predict_combinatorial_ko usage

100%

gene_pairs parameter

100%

synergy_threshold parameter

100%

Negative controls included

16%

33%

Gene vocabulary check

40%

Cell type distribution check

37%

Cross-model validation

30%

Standard cell type name

100%

predict_dose_response usage

18%

export_validation_guide usage

Valid perturbation_type

100%

Without context: $0.3665 · 1m 49s · 13 turns · 18 in / 7,423 out tokens

With context: $1.1915 · 3m 43s · 32 turns · 262 in / 11,738 out tokens

Evaluated: 1 day ago
Agent: Claude Code

Table of Contents

Oncology Target Prioritization Report Transcriptomics Pipeline for Cardiac Gene Perturbation Study Synergy Screening for Combination Immunotherapy Targets

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.