Garden announced BLOOM (Branching Lookup Optimized for Organic Molecules), a Markush structure search engine that gives AI drug-design teams near-instant verification of small-molecule IP landscapes, so researchers can iterate on candidates with legal certainty built into the loop. As AI models generate molecules faster than ever, the bottleneck has become diligence: knowing what’s already “covered” before committing lab time and budget. BLOOM removes that bottleneck.
BLOOM uses a graph-based, agentic traversal to compare Markush queries against millions of SMILES strings, short-circuiting invalid candidates via local atom and bond features. The system then delivers color-coded mapping that confirms atom- and bond-level compliance, turning verification from a manual chore into a built-in step.
In benchmark testing, BLOOM delivered an average 32.44× speed improvement over a standard core-extraction string search (0.047 ms vs. 1.491 ms per comparison) while finding correct matches that string methods missed, including a designed single-hit query across a multi-million-record corpus. For IP-aware design loops, that means go/no-go decisions during ideation, not weeks later.
Also Read: HistoIndex Advances Fibrosis Assessment with FibroSIGHT™ Plus
BLOOM also reduces false positives that plague legacy approaches, which often fail on nuanced bond counts and positioning, saving teams from time-consuming, atom-by-atom checks. With BLOOM, verification is automatic and scalable.
The engine is integrated with Garden’s patent database: every SMILES match links to underlying patent records, and Garden’s AI agent can summarize, compare, and help prune result sets. BLOOM supports workflows from rapid novelty triage to freedom-to-operate analysis directly alongside model-driven design.
“AI can propose thousands of viable chemistries in minutes. BLOOM closes the loop by telling you what’s already fenced off instantly,” said Adi Sidapara, founder and CEO of Garden. “You get IP-aware exploration without slowing down discovery.”
“Small changes around an R-group can define patentability,” added Kavin Sivakumar, Ph.D., Founding ML Researcher at Garden. “BLOOM’s graph reasoning captures those subtleties at speed, so IP checks no longer throttle design.”
Source: PRNewswire
