Curated AI beats frontier LLMs at pharma asset discovery

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Kidziński, Łukasz, Thomas, Kevin
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910194526584832
author Kidziński, Łukasz
Thomas, Kevin
author_facet Kidziński, Łukasz
Thomas, Kevin
contents General-purpose LLMs with web search are increasingly used to scout the competitive landscape of pharmaceutical pipelines. We benchmark Gosset -- an AI platform with a chat interface backed by curated target-, modality-, and indication-level drug-asset annotations -- against four frontier systems with web access (Claude Opus 4.7, GPT 5.5, Gemini 3.1 Pro, Perplexity sonar-pro) on ten niche oncology/immunology targets where most of the pipeline lives in the long tail of preclinical and Asian-developed assets. All five systems receive the same natural-language query and the same JSON output schema. Across 10 targets Gosset returns 3.2x more verified drugs per query than the best frontier system, at perfect precision and 100% recall against the cross-system union of verified drugs. The same curated index is exposed as a Gosset MCP server that any frontier model can call as a tool, suggesting that each of these systems can close most of the recall gap by swapping generic web search for a curated index behind the same chat interface.
format Preprint
id arxiv_https___arxiv_org_abs_2605_04908
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Curated AI beats frontier LLMs at pharma asset discovery
Kidziński, Łukasz
Thomas, Kevin
Artificial Intelligence
Quantitative Methods
H.3.3; I.2.7; J.3
General-purpose LLMs with web search are increasingly used to scout the competitive landscape of pharmaceutical pipelines. We benchmark Gosset -- an AI platform with a chat interface backed by curated target-, modality-, and indication-level drug-asset annotations -- against four frontier systems with web access (Claude Opus 4.7, GPT 5.5, Gemini 3.1 Pro, Perplexity sonar-pro) on ten niche oncology/immunology targets where most of the pipeline lives in the long tail of preclinical and Asian-developed assets. All five systems receive the same natural-language query and the same JSON output schema. Across 10 targets Gosset returns 3.2x more verified drugs per query than the best frontier system, at perfect precision and 100% recall against the cross-system union of verified drugs. The same curated index is exposed as a Gosset MCP server that any frontier model can call as a tool, suggesting that each of these systems can close most of the recall gap by swapping generic web search for a curated index behind the same chat interface.
title Curated AI beats frontier LLMs at pharma asset discovery
topic Artificial Intelligence
Quantitative Methods
H.3.3; I.2.7; J.3
url https://arxiv.org/abs/2605.04908