Saddle Point Science
The advanced statistical analysis platform for modern precision medicine
Saddle Point Science develops cutting-edge software tailored to overcome the many challenges posed by modern, complex biomedical data. Building on our proven scientific track record, we specialise in creating novel methods that extract as much information as possible from your data, to deliver robust, reliable inferences and prediction models. Our software suite offers a range of inference and regression pipelines and includes multiple models for data visualisation, data preprocessing and manipulation, validation, and supports fully automated reporting.
Unlock the Potential of Your Biomedical Data
Saddle Point Science provides advanced statistical data analysis to uncover meaningful patterns in complex biomedical datasets. From patient stratification and treatment response to risk prediction and biomarker discovery, we help you turn complex data into actionable insights that support scientific discovery and better-informed decisions.
Have a dataset you want to explore? Let's discover what your data can reveal.
What we help you achieve
Our Solutions
1
spsSIGNATURE
Generate individual-level risk and treatment response signatures from high-dimensional biomedical data. Designed to address overfitting and statistical bias, spsSIGNATURE incorporates advanced statistical methodology into an automated analytical pipeline.
Key capabilities
-
Personalised risk prediction
-
Treatment response prediction
-
Time-to-event, ordinal, and real-valued response clinical outcome
-
Probabilistic variable selection
-
Overfitting bias correction and statistical regularisation
-
Correction for informative data missingness
-
Shadow analysis and automated reporting
2
spsMOSAICS
Not all patients respond to treatment in the same way. Relevant differences between patients may exist even when measured characteristics do not clearly distinguish them. spsMOSAICS identifies latent patient subgroups and assesses differences in risk, survival and treatment associations.
Key capabilities
-
Bayesian latent patient subgroup discovery
-
Quantification of subgroup membership probabilities
-
Identification of potential treatment responder groups
-
Patient stratification and subgroup characterisation
-
Exploration of potential stratifying biomarkers
-
Analysis of multiple risks and competing-risk effects
3
BFI
Combining data across hospitals and research centres can be difficult because of privacy requirements, consent limitations and organisational barriers. BFI (Bayesian Federated Inference) enables statistical information from separate datasets to be combined without requiring the underlying patient-level data to be pooled.
Key capabilities
-
Accurate statistical inference across distributed datasets
-
Local analysis at participating centres
-
Multi-centre research without centralising patient-level data
-
Analysis designed for challenging biomedical research settings
-
Combined statistical insights under applicable modelling assumptions
Who we currently support
Pharmaceutical
companies
Biotechnology companies
Academic Researchers
Clinical trial organizations
Partners
.jpg){kind=logo}
Publications & Success Stories
Scientific publications
Latest news
