Loading BioAtlas...
Loading BioAtlas...
مقایسه شاخ‌به‌شاخ مدل‌های زیستی، موتورهای داکینگ و پایپ‌لاین‌های محاسباتی. فیلتر بر اساس مدل دسترسی، مصرف پردازشی و دسته‌بندی جهت اخذ تصمیم دقیق در کمتر از یک دقیقه.
| مشخصه | ＋ افزودن ابزار | ||
|---|---|---|---|
| Category | Genomics & Bioinformatics | Genomics & Sequence Modeling | — |
| Subcategory | Genomic Foundation Models | Bi-directional Mamba Genomic Models | — |
| Company | Arc Institute & Stanford University | Stanford University & Princeton University | — |
| Access mode | Open Source | Open Source | — |
| Licence conditions | Apache License 2.0 | Apache 2.0 | — |
| Compute requirements | 1x A100/H100 GPU (80GB) for inference; multi-node cluster for fine-tuning | 1x GPU (>= 16GB VRAM) | — |
| Biological application | Synthetic biology, CRISPR system design, regulatory sequence engineering, and whole-genome generation | Long-range gene regulation, promoter-enhancer mapping, and variant impact scoring | — |
| Research workflow | Input DNA sequence (up to 131kb) -> predict fitness effects or generate de novo genomic sequences | Genomic DNA sequence -> predict regulatory activity with RC equivariance | — |
| Evidence level | Peer-reviewed publication (Science 2024) | Peer-reviewed research (ICML 2024) | — |
| Commercial eligibility | Open weights and open-source code | Open source release | — |
| Price | Free Open Source | Free Open Source | — |
| Availability ⓘ | Available on GitHub, Hugging Face, and Together AI | Available on GitHub and Hugging Face | — |
| Features | 7 billion parameters trained on 300 billion nucleotides across prokaryotic and viral diversity, 131,072 token context window powered by the StripedHyena architecture, Predicts mutation effects across protein coding and non-coding regulatory elements, Generates novel, functional CRISPR-Cas molecular complexes and synthetic transposable elements | Bi-directional state space (Mamba) architecture overcoming causal autoregressive limitations, Guaranteed reverse-complement (RC) equivariance essential for double-stranded DNA modeling, Sub-quadratic scaling allowing 131k base pair context on single standard GPUs, Superior performance on variant effect and chromatin accessibility benchmarks | — |
| Entry type | Open Source Model | AI Model | — |
| AI role | Whole-Genome Sequence Generation & Fitness Prediction | Equivariant State Space Modeling | — |
| Input data | ثبت‌نشده | ثبت‌نشده | — |
| Output data | ثبت‌نشده | ثبت‌نشده | — |
| Validation | ثبت‌نشده | ثبت‌نشده | — |
| Type | Long-context genomic foundation model | Bi-directional Mamba genomic model | — |
| Intended use | ثبت‌نشده | ثبت‌نشده | — |
| Compatibility | ثبت‌نشده | ثبت‌نشده | — |
| Manufacturer | Arc Institute, Stanford University & Together AI | Stanford University & Princeton University | — |
| Integration evidence | https://github.com/evo-design/evo | https://github.com/kuleshov-group/caduceus | — |
| Laboratory handoff | Generated operons and CRISPR constructs synthesized via commercial gene synthesis for functional validation | Informs non-coding variant prioritization for CRISPR functional genomics | — |