Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -7,10 +7,10 @@ license: mit
|
|
| 7 |
> **A growing benchmark for evaluating LLM agents on complex bioinformatics workflows.**
|
| 8 |
|
| 9 |
[](https://huggingface.co/datasets/lingzhi227/Extended-BioAgentBench)
|
| 10 |
-
[](https://github.com/bioagent-bench/bioagent-bench)
|
| 12 |
|
| 13 |
-
Building on [BioAgentBench](https://arxiv.org/abs/2601.21800) (10 tasks), this benchmark adds **
|
| 14 |
|
| 15 |
**This benchmark is actively growing** — new tasks are added continuously to cover more domains, increase complexity, and push the boundaries of what AI agents can do in bioinformatics.
|
| 16 |
|
|
@@ -21,7 +21,7 @@ Building on [BioAgentBench](https://arxiv.org/abs/2601.21800) (10 tasks), this b
|
|
| 21 |
- **Domain-specific traps**: Tasks include steps where default parameters silently produce wrong results (e.g., Tn5 shift correction in ATAC-seq, Medaka model selection for Nanopore)
|
| 22 |
- **Real public data**: Every task uses published datasets with ground truth generated by validated reference pipelines
|
| 23 |
|
| 24 |
-
## Tasks (
|
| 25 |
|
| 26 |
| # | Task ID | Name |
|
| 27 |
|---|---------|------|
|
|
@@ -54,6 +54,7 @@ Building on [BioAgentBench](https://arxiv.org/abs/2601.21800) (10 tasks), this b
|
|
| 54 |
| 37 | `cutandrun` | CUT&RUN Epigenomic Profiling |
|
| 55 |
| 38 | `scrna-full-pipeline` | Single-Cell RNA-seq Full Pipeline: Multi-Quantifier Ana |
|
| 56 |
| 39 | `crispr-screen` | CRISPR Screen Analysis: Drug Sensitivity Gene Discovery |
|
|
|
|
| 57 |
|
| 58 |
## Quick start
|
| 59 |
|
|
|
|
| 7 |
> **A growing benchmark for evaluating LLM agents on complex bioinformatics workflows.**
|
| 8 |
|
| 9 |
[](https://huggingface.co/datasets/lingzhi227/Extended-BioAgentBench)
|
| 10 |
+
[]()
|
| 11 |
[](https://github.com/bioagent-bench/bioagent-bench)
|
| 12 |
|
| 13 |
+
Building on [BioAgentBench](https://arxiv.org/abs/2601.21800) (10 tasks), this benchmark adds **30 new tasks** that test LLM agents on increasingly complex, multi-tool bioinformatics pipelines. Tasks span 6 domains and range from simple linear workflows to depth-8 diamond DAGs with 16+ CLI tools.
|
| 14 |
|
| 15 |
**This benchmark is actively growing** — new tasks are added continuously to cover more domains, increase complexity, and push the boundaries of what AI agents can do in bioinformatics.
|
| 16 |
|
|
|
|
| 21 |
- **Domain-specific traps**: Tasks include steps where default parameters silently produce wrong results (e.g., Tn5 shift correction in ATAC-seq, Medaka model selection for Nanopore)
|
| 22 |
- **Real public data**: Every task uses published datasets with ground truth generated by validated reference pipelines
|
| 23 |
|
| 24 |
+
## Tasks (30 total)
|
| 25 |
|
| 26 |
| # | Task ID | Name |
|
| 27 |
|---|---------|------|
|
|
|
|
| 54 |
| 37 | `cutandrun` | CUT&RUN Epigenomic Profiling |
|
| 55 |
| 38 | `scrna-full-pipeline` | Single-Cell RNA-seq Full Pipeline: Multi-Quantifier Ana |
|
| 56 |
| 39 | `crispr-screen` | CRISPR Screen Analysis: Drug Sensitivity Gene Discovery |
|
| 57 |
+
| 40 | `amplicon-microbiome` | 16S Amplicon Microbiome: Community Profiling and Functi |
|
| 58 |
|
| 59 |
## Quick start
|
| 60 |
|