ForecomAI (@ForecomAI) header
ForecomAI (@ForecomAI) profile picture

Bespoke AI for biosciences: cell imaging, multi-modal data fusion, bioprocess optimisation and real-world evidence. @forecomai.bsky.social

United Kingdom
Joined March 2023
Rabies virus produces no cytopathic effects, so this AI reads viral antigen fluorescence instead. MambaOut: 99.6% well accuracy, and expert-matched titres on the WHO/WOAH assay without retraining. nature.com/articles/s41598-0…
1
23
Work by ForecomAI CTO Prof. Miroslaw Bober with APHA and the University of Surrey. The wider aim: resolving weak biological signals robustly enough to make cell-based assays quantitative.
17
ForecomAI retweeted
How long should you treat venous thromboembolism (VTE)? Not simply 3–6 months or indefinitely. The challenge is balancing the risks of recurrent clots and bleeding. Our 2 new JTH papers provide complementary dynamic risk prediction tools supporting personalised treatment decision
3
4
396
How long should you treat venous thromboembolism (VTE)? There's no fixed answer — it's a balance between preventing another clot and avoiding the risk of bleeding. Our two new studies offer simple tools to help personalise that decision for each patient. Thread
1
18
📢 Excited to share that our joint work with CVSSP (University of Surrey) will be presented at CVPR 2025 this week in Nashville. @UniOfSurrey @PeopleCentredAI #CVPR2025 #SSL #ComputationalBiology #MorphologicalProfiling #BioAI
1
94
We introduce TRex — a two-stage framework that adapts static SSL features to biologically meaningful tasks like MoA classification. It doubles mAP (0.15 → 0.32), improves compound recognition and generalisation — without retraining large models.
1
58
PAI's Professor Miroslaw Bober will be presenting the work @UniOfSurrey @PeopleCentredAI
30
Why do pre-o3 LLMs struggle with generalization tasks like @arcprize? It's not what you might think. OpenAI o3 shattered the ARC-AGI benchmark. But the hardest puzzles didn’t stump it because of reasoning, and this has implications for the benchmark as a whole. Analysis below🧵
18
65
14
645
206,791
.@PeopleCentredAI unveils a cutting-edge #AI system that can accurately and quickly identify protein patterns in cells. The new system has the potential to speed progress in #cancerresearch ow.ly/yAm950Ok3Wl #ProudtobeSurrey
6
4
4
1,850
AI developed in the UK is the world leader in identifying the location and expression of proteins surrey.ac.uk/news/ai-develop…
1
64
Join the State of Precision Medicine webinar, where physicians, scientists, directors, and patient advocates will discuss the current state and future of the field. Speakers include Professor Mathias Uhlén who will discuss related advances of multiomics. proteinatlas.org/news/2023-0…
1
3
5
706
Really proud to be published in a Nature Portfolio journal for the first time! We set a new SOTA for single-cell protein localisation on the @ProteinAtlas, building on our work in the 2nd HPA Kaggle comp. nature.com/articles/s42003-0… @ForecomAI @cvssp_research @d_minskiy
3
5
1
21
7,233
RT Single-cell subcellular protein localisation using novel ensembles of diverse deep architectures dlvr.it/SnZYht
1
31
The Hybrid subCellular Protein Localiser (HCPL) improves single-cell classification for protein localization, allowing for large-scale data annotation through deep-learning architecture. nature.com/articles/s42003-0…
2
6
687
We are excited to share that our work on an advanced AI system for single-cell image analysis of protein localisation has been published in a Nature portfolio journal rdcu.be/dbsFC #SingleCell #ArtificialIntelligence #MachineLearning #ImageAnalysis #Microscopy
1
1
177
This technique can rapidly identify the subcellular localisation of proteins of interest in images, streamlining high throughput screening for various applications including the study of tumour heterogeneity, identification of disease and toxicity biomarkers, and drug discovery.
1
22
This was a successful partnership between @cvssp_research and @ForecomAI
13