{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:QOQAVYY43YPYNKHKKZWIYBHJLV","short_pith_number":"pith:QOQAVYY4","schema_version":"1.0","canonical_sha256":"83a00ae31cde1f86a8ea566c8c04e95d654531184adaea6006f9a4ffe0cafb51","source":{"kind":"arxiv","id":"2605.17261","version":1},"attestation_state":"computed","paper":{"title":"Unlocking Biological Workflows for Robust Protein-Text Question Answering: A Dual-Dimensional RAG Framework","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"2D-ProteinRAG embeds LLMs in BLAST workflows with dual filtering to handle novel proteins in question answering","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Chen Huang, Duanyu Feng, Li Ding, See-kiong Ng, Wenqiang Lei, Yang Li, Yangshuai Wang","submitted_at":"2026-05-17T05:03:24Z","abstract_excerpt":"Protein-Text Question Answering (QA) is crucial for interpreting biological sequences through natural language. The integration of Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) that efficiently leverages biological databases and facilitates reasoning offers a potent approach for it. However, constrained by the standard RAG pipeline, these models often rely on curated, static datasets instead of expert-proven biological workflows, lacking the fine-grained information processing and struggling to generalize to novel (OOD) proteins. To bridge this gap, we propose 2D-Prote"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.17261","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2026-05-17T05:03:24Z","cross_cats_sorted":[],"title_canon_sha256":"c31592c0b5e6642281f93c4b001ef43594e0eac7f39013260bcc93ae2580591f","abstract_canon_sha256":"b057d46bd46226285775af30661f89963b9652bb9a3ef8661d75363266f42e15"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:03:48.300619Z","signature_b64":"hxvBbZ3d8KMNDAzvsHn6Z11mbdAlhbYCBELa1GZDEFe797Hk26PzE6o/IJ4jh6yMA/0iL8PfKhT+4icfxxtPBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"83a00ae31cde1f86a8ea566c8c04e95d654531184adaea6006f9a4ffe0cafb51","last_reissued_at":"2026-05-20T00:03:48.299779Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:03:48.299779Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Unlocking Biological Workflows for Robust Protein-Text Question Answering: A Dual-Dimensional RAG Framework","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"2D-ProteinRAG embeds LLMs in BLAST workflows with dual filtering to handle novel proteins in question answering","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Chen Huang, Duanyu Feng, Li Ding, See-kiong Ng, Wenqiang Lei, Yang Li, Yangshuai Wang","submitted_at":"2026-05-17T05:03:24Z","abstract_excerpt":"Protein-Text Question Answering (QA) is crucial for interpreting biological sequences through natural language. The integration of Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) that efficiently leverages biological databases and facilitates reasoning offers a potent approach for it. However, constrained by the standard RAG pipeline, these models often rely on curated, static datasets instead of expert-proven biological workflows, lacking the fine-grained information processing and struggling to generalize to novel (OOD) proteins. To bridge this gap, we propose 2D-Prote"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Extensive evaluations on both In-Distribution and diverse biological OOD benchmarks demonstrate that 2D-ProteinRAG consistently achieves state-of-the-art performance, outperforming fine-tuned baselines and other RAG methods.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the proposed horizontal fine-grained attribute alignment and vertical homology-based semantic denoising steps, when applied after BLAST retrieval, will reliably extract high-quality information from noisy contexts and generalize to novel proteins without introducing new errors.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"2D-ProteinRAG is a dual-dimensional RAG framework that incorporates BLAST workflows plus horizontal attribute alignment and vertical homology denoising to improve protein-text QA on both in-distribution and out-of-distribution cases.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"2D-ProteinRAG embeds LLMs in BLAST workflows with dual filtering to handle novel proteins in question answering","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"5f6ddd98eff7249e03553299847d84de8db40366a40456015a569ce169c0cbd4"},"source":{"id":"2605.17261","kind":"arxiv","version":1},"verdict":{"id":"85851c50-52de-4533-b42f-4428cd84c4e6","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T23:27:10.897126Z","strongest_claim":"Extensive evaluations on both In-Distribution and diverse biological OOD benchmarks demonstrate that 2D-ProteinRAG consistently achieves state-of-the-art performance, outperforming fine-tuned baselines and other RAG methods.","one_line_summary":"2D-ProteinRAG is a dual-dimensional RAG framework that incorporates BLAST workflows plus horizontal attribute alignment and vertical homology denoising to improve protein-text QA on both in-distribution and out-of-distribution cases.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the proposed horizontal fine-grained attribute alignment and vertical homology-based semantic denoising steps, when applied after BLAST retrieval, will reliably extract high-quality information from noisy contexts and generalize to novel proteins without introducing new errors.","pith_extraction_headline":"2D-ProteinRAG embeds LLMs in BLAST workflows with dual filtering to handle novel proteins in question answering"},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.17261/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T23:31:20.257647Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T23:31:05.313423Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T22:01:57.849591Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.784016Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"7e8f7d884be65662258863fc3d81c503b781f7b6afa997208b009b5e9129ca8b"},"references":{"count":38,"sample":[{"doi":"","year":1990,"title":"S F Altschul, W Gish, W Miller, E W Myers, and D J Lipman. 1990. Basic local alignment search tool.J. Mol. Biol.215, 3 (Oct. 1990), 403–410","work_id":"813b4ae7-7b1d-4ebf-8d62-7ff7859783c9","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities","work_id":"008df105-2fdd-45d8-857a-8e35868aecb6","ref_index":2,"cited_arxiv_id":"2507.06261","is_internal_anchor":true},{"doi":"10.1093/nar/gkae1010","year":2024,"title":"The UniProt Consortium. 2024. UniProt: the Universal Pro- tein Knowledgebase in 2025.Nucleic Acids Research53, D1 (11 2024), D609–D617. arXiv:https://academic.oup.com/nar/article- pdf/53/D1/D609/60719","work_id":"dbdc78a3-ee26-47ca-892c-66b69b43dbae","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2000,"title":"D Devos and A Valencia. 2000. Practical limits of function prediction.Proteins 41, 1 (Oct. 2000), 98–107","work_id":"7977ae4f-abfc-4c04-8527-e39412efcaab","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"arXiv preprint arXiv:2501.10282 (2025)","work_id":"82aa0ed8-bab7-4c7b-9892-2eac426c3055","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":38,"snapshot_sha256":"a9f8197e1000ef29d95947269b20f08506372b4942e6472771458152f24c75e2","internal_anchors":4},"formal_canon":{"evidence_count":1,"snapshot_sha256":"2280aeb9f0ab83793726399ba71a2669d64c9c4e7a04468bb425bf121b8aea47"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2605.17261","created_at":"2026-05-20T00:03:48.299910+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.17261v1","created_at":"2026-05-20T00:03:48.299910+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17261","created_at":"2026-05-20T00:03:48.299910+00:00"},{"alias_kind":"pith_short_12","alias_value":"QOQAVYY43YPY","created_at":"2026-05-20T00:03:48.299910+00:00"},{"alias_kind":"pith_short_16","alias_value":"QOQAVYY43YPYNKHK","created_at":"2026-05-20T00:03:48.299910+00:00"},{"alias_kind":"pith_short_8","alias_value":"QOQAVYY4","created_at":"2026-05-20T00:03:48.299910+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":1,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV","json":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV.json","graph_json":"https://pith.science/api/pith-number/QOQAVYY43YPYNKHKKZWIYBHJLV/graph.json","events_json":"https://pith.science/api/pith-number/QOQAVYY43YPYNKHKKZWIYBHJLV/events.json","paper":"https://pith.science/paper/QOQAVYY4"},"agent_actions":{"view_html":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV","download_json":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV.json","view_paper":"https://pith.science/paper/QOQAVYY4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.17261&json=true","fetch_graph":"https://pith.science/api/pith-number/QOQAVYY43YPYNKHKKZWIYBHJLV/graph.json","fetch_events":"https://pith.science/api/pith-number/QOQAVYY43YPYNKHKKZWIYBHJLV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV/action/storage_attestation","attest_author":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV/action/author_attestation","sign_citation":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV/action/citation_signature","submit_replication":"https://pith.science/pith/QOQAVYY43YPYNKHKKZWIYBHJLV/action/replication_record"}},"created_at":"2026-05-20T00:03:48.299910+00:00","updated_at":"2026-05-20T00:03:48.299910+00:00"}