Loading BioAtlas...
Loading BioAtlas...
将生物大模型、对接计算引擎与组学管线并排横向对比。按获取方式、硬件开销及细分领域灵活筛选，在一分钟内得出客观可辩护的技术选型方案。
| 技术参数指标 | ＋ 添加工具 | |||||
|---|---|---|---|---|---|---|
| Category | Genomics & Sequence Modeling | Genomics & Bioinformatics | Genomics & Bioinformatics | Genomics & Bioinformatics | Genomics & Bioinformatics | — |
| Subcategory | 3D Variant Pathogenicity | Regulatory Genomics | Conventional or Statistical Baselines | Research AI Methods | Research AI Methods | — |
| Company | Illumina Artificial Intelligence Lab | Google DeepMind | Research / method authors | Research / method authors | Research / method authors | — |
| Access mode | Open Access for Research | Research code or publication | Confirm distribution and exact release | Confirm distribution and exact release | Confirm distribution and exact release | — |
| Licence conditions | Academic non-commercial use | Check code and model weights separately | NOT_INDEPENDENTLY_AUDITED | NOT_INDEPENDENTLY_AUDITED | NOT_INDEPENDENTLY_AUDITED | — |
| Compute requirements | Pre-computed genome-wide lookup table / GPU inference | Check selected checkpoint and runtime; no GPU catalog recommendation | Version and data dependent | Version and data dependent | Version and data dependent | — |
| Biological application | Rare disease genetic diagnosis, carrier screening, and drug target discovery | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | — |
| Research workflow | Patient genomic variant -> query 3D pathogenicity score | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | — |
| Evidence level | Peer-reviewed research (Science 2023) | Peer-reviewed primary method | PEER_REVIEWED_METHOD_OR_CONFERENCE | PEER_REVIEWED_METHOD_OR_CONFERENCE | PEER_REVIEWED_METHOD_OR_CONFERENCE | — |
| Commercial eligibility | Open access for non-commercial research; Illumina platform integration | Unknown, requires audit | NOT_INDEPENDENTLY_AUDITED | NOT_INDEPENDENTLY_AUDITED | NOT_INDEPENDENTLY_AUDITED | — |
| Price | Free for academic research | Check selected model licence | Verify access and licence | Verify access and licence | Verify access and licence | — |
| Availability ⓘ | Available via Illumina GitHub and Connected Analytics | Public research method/code | Method paper or official documentation | Method paper or official documentation | Method paper or official documentation | — |
| Features | Trained on benign variants identified across 233 diverse primate species, Directly incorporates 3D structural protein atomic coordinate representations, Outperforms previous variant classification tools on ClinVar and UK Biobank cohorts, Classifies variants of unknown significance (VUS) in rare genetic disorders | Predicts genomic regulatory tracks and predicted functional effects of noncoding DNA variants., Genomic regulatory track, sequence-variant contrast | bcftools scientific method, Reference genome with aligned short-read input and correct ploidy assumptions, VCF/BCF variant calls and variants QC operations | CADD scientific method, Human genomic variant coordinates mapped to correct reference assembly, Combined annotation-dependent depletion scores | EVE scientific method, Protein family homolog alignment and compatible human missense variants, Evolutionary deep generative variant effect estimate | — |
| Entry type | AI Model & Database | Research AI model / method | Classical/statistical scientific method | Source-scoped research method | Source-scoped research method | — |
| AI role | 3D Structural Variant Classification | Published biomedical AI method | NON_AI_STATISTICAL | ENSEMBLE_ML | DEEP_LEARNING | — |
| Input data | 未明确记录 | DNA sequence interval; alleles and tissue/assay ontology | Reference genome with aligned short-read input and correct ploidy assumptions | Human genomic variant coordinates mapped to correct reference assembly | Protein family homolog alignment and compatible human missense variants | — |
| Output data | 未明确记录 | Genomic regulatory track; sequence-variant contrast | VCF/BCF variant calls and variants QC operations | Combined annotation-dependent depletion scores | Evolutionary deep generative variant effect estimate | — |
| Validation | 未明确记录 | Requires independent dataset-specific evaluation | Published method scope; no independent BioAtlas execution or benchmark | Published method scope; no independent BioAtlas execution or benchmark | Published method scope; no independent BioAtlas execution or benchmark | — |
| Type | 3D deep learning variant pathogenicity classifier | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | — |
| Intended use | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | — |
| Compatibility | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | — |
| Manufacturer | Illumina, Inc. | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | — |
| Integration evidence | https://github.com/illumina/PrimateAI-3D | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | — |
| Laboratory handoff | Directly used in clinical genetics diagnostic reporting | 未明确记录 | 未明确记录 | 未明确记录 | 未明确记录 | — |