Interrogate biological systems
We use CRISPR and pharmacological perturbations to reveal genes, pathways and interactions that determine cancer-cell fitness and drug response.
Predictive Perturbational Biology
From cancer vulnerabilities to therapeutic targets
We combine functional genomics, multi-omics and machine learning to understand how cancer cells respond to genetic and pharmacological perturbations, and to predict vulnerabilities that can be exploited therapeutically.
Functional genomics · Cancer dependencies · Machine learningHuman Technopole · Milan
Our approach
Perturb, Measure, Predict
We use systematic experimental perturbations to learn the rules governing cancer-cell behaviour.
Resolve cellular response
We integrate molecular profiles across time, scale and context to distinguish specific dependency responses from shared cellular programmes.
Learn transferable rules
We develop statistical and machine-learning methods that generalise from existing screens to predict vulnerabilities in unseen contexts.
Research programmes
What we are working on now
A common question connects our projects: can we move from cataloguing cancer vulnerabilities to understanding and predicting them?
01 · DepSHOCK
What happens after acute inactivation of a genetic dependency?
Longitudinal perturbational profiling and unsupervised machine learning to characterise cancer dependency shock (DepSHOCK), uncover the pathways sustaining genetic dependencies, and build models that predict dependency from DepSHOCK response signatures.
02 · CRISPR-VUS / DAM
Turning cancer variation into functional evidence
Large-scale functional maps that connect variants of uncertain significance to cancer dependencies, molecular context and therapeutic hypotheses.
03 · Genetic interactions
Predicting synthetic lethalities beyond measured pairs
Uncertainty-aware models integrating genome-wide single-gene screens with targeted combinatorial perturbations.
04 · Drug sensitivity & recovery
Learning vulnerabilities created by adaptation
Mapping how cancer cells acquire, maintain and reverse drug-resistant states to expose new therapeutic opportunities.
Every perturbation teaches us. Every model helps us choose the next experiment to perform.
IorioLab · Computational Biology Research Centre · Human Technopole
Open science
Tools, data and resources
We are committed to open and reproducible science, sharing our methods, data and computable scientific manuscripts so others can reproduce, scrutinise and build on our work. Explore our dashboard of software tools and computable scientific manuscripts.
Software
CRISPRcleanR
Correction of gene-independent responses in CRISPR-Cas9 screens.
Method
ADaM
Identification of context-specific cancer dependencies from functional-genomic screens.
Resource
CRISPR-VUS / DAM
Functional interpretation of cancer genetic variation at scale.
Programme
PRECISE / EuroDepMap
Collaborative efforts toward predictive functional genomics and cancer dependency mapping.
The lab
One group. Two arms. Algorithms and experiments
We bring computational scientists, molecular biologists and genome engineers together around the same biological questions. Computation shapes the experiments; experiments reshape the models.
Latest
From the lab
Publications, projects, events and opportunities.
Research
CRISPR-VUS: functional interpretation of cancer variation at scale
Project update and associated resources.
Consortium
PRECISE: toward predictive models of cancer vulnerabilities
Collaborative programme connecting perturbational data and predictive modelling.
Opportunities
Interested in predictive perturbational biology?
See current opportunities and information for prospective lab members.