Spaces:
Sleeping
Sleeping
Merge pull request #8 from KevinIsInCoding/docs/prfaq-compound-discovery
Browse files
PRFAQ.md
CHANGED
|
@@ -10,7 +10,7 @@
|
|
| 10 |
|
| 11 |
*New AI-powered tool synthesizes 500+ peer-reviewed ALS papers and live clinical trial data into a single research landscape report, grounded in a biomedical knowledge graph*
|
| 12 |
|
| 13 |
-
**New
|
| 14 |
|
| 15 |
ALS is one of the most aggressive and least understood neurodegenerative diseases. With over 100 active clinical trials testing targets ranging from SOD1 gene silencing to neuroinflammation modulators, physicians face a mounting challenge: how do you evaluate a trial's scientific credibility when the evidence base spans thousands of papers across genetics, cell biology, and clinical pharmacology?
|
| 16 |
|
|
@@ -20,6 +20,8 @@ Candle-Fire changes this. A physician types a natural-language question — "Wha
|
|
| 20 |
|
| 21 |
The system combines two layers of intelligence. A biomedical knowledge graph connects genes, proteins, compounds, pathways, and clinical trials extracted from 500 curated ALS papers, enabling the system to surface related biology that a keyword search would miss. A vector retrieval layer finds the most relevant passages from those papers and weights them by citation count, surfacing high-impact evidence over peripheral findings. Claude Sonnet synthesizes both into a physician-readable report, streamed in real time.
|
| 22 |
|
|
|
|
|
|
|
| 23 |
"ALS research moves fast, but the tools physicians use to stay current haven't kept up," said the Candle-Fire team. "We built this because the gap between what the research community knows and what a treating neurologist can practically access in a clinic visit is too wide."
|
| 24 |
|
| 25 |
Candle-Fire is available as a web app on HuggingFace Spaces. It is free to use and open-source under the MIT License.
|
|
|
|
| 10 |
|
| 11 |
*New AI-powered tool synthesizes 500+ peer-reviewed ALS papers and live clinical trial data into a single research landscape report, grounded in a biomedical knowledge graph*
|
| 12 |
|
| 13 |
+
**New Hyde Park, NY** — 12/30/2026 Today, the team behind Beacon announced the launch of Candle-Fire, a research intelligence platform that helps physicians evaluating ALS (amyotrophic lateral sclerosis) clinical trials understand the full scientific context behind a trial's mechanism; to help researcher to understand the disease pathology and existing experiment in order to better locate disease target.
|
| 14 |
|
| 15 |
ALS is one of the most aggressive and least understood neurodegenerative diseases. With over 100 active clinical trials testing targets ranging from SOD1 gene silencing to neuroinflammation modulators, physicians face a mounting challenge: how do you evaluate a trial's scientific credibility when the evidence base spans thousands of papers across genetics, cell biology, and clinical pharmacology?
|
| 16 |
|
|
|
|
| 20 |
|
| 21 |
The system combines two layers of intelligence. A biomedical knowledge graph connects genes, proteins, compounds, pathways, and clinical trials extracted from 500 curated ALS papers, enabling the system to surface related biology that a keyword search would miss. A vector retrieval layer finds the most relevant passages from those papers and weights them by citation count, surfacing high-impact evidence over peripheral findings. Claude Sonnet synthesizes both into a physician-readable report, streamed in real time.
|
| 22 |
|
| 23 |
+
Beyond individual patient care, the team envisions Candle-Fire playing a longer-term role in accelerating compound discovery. When researchers can rapidly map which biological pathways remain poorly targeted, which compound classes have already been tested and failed, and where the strongest mechanistic evidence clusters, they gain a clearer picture of the white space worth pursuing. A system that continuously synthesizes the growing ALS literature — tracking which gene-compound relationships are well-evidenced and which are hypothetical — can serve as an early signal layer for drug discovery teams deciding where to focus preclinical investment. As the knowledge graph matures and the corpus expands, Candle-Fire's relationship network between compounds, targets, and mechanisms has the potential to surface novel combination hypotheses that no single research team could derive from manual review alone.
|
| 24 |
+
|
| 25 |
"ALS research moves fast, but the tools physicians use to stay current haven't kept up," said the Candle-Fire team. "We built this because the gap between what the research community knows and what a treating neurologist can practically access in a clinic visit is too wide."
|
| 26 |
|
| 27 |
Candle-Fire is available as a web app on HuggingFace Spaces. It is free to use and open-source under the MIT License.
|