ALaSCA: A novel in silico
simulation platform to untangle
biological pathway
mechanisms, with a case study
in Type 1 Diabetes progression
Dr Raminderpal Singh
CEO, Incubate bio
raminderpal@incubate.bio
30 March 2023

Presentation outline
1. Introduction to our technology (ALaSCA)
2. Demonstration on Type 1 Diabetes
For extended discussion:
2
https://www.biorxiv.org/content/10.1101/2023.03.16.532913v1
POSTER #11

Introduction to causal analysis
What is causal inference?
○ ALaSCA is based on formal Pearlian causal inference (PCI) techniques. Using PCI, we are able to
quantify causal relationships using an array of different data types that span the hierarchy of
biological organisation from gene to phenotype.
○ ALaSCA not only takes into account the causal relationships and their immediate surrounding
network, the causal analyses and counterfactual simulation are performed in context of the
entire biological mechanism.
What is counterfactual simulation?
○ Counterfactual, or ‘contrary-to-the-facts’ simulation, is a conviction based upon the human
inclination to envision alternate outcomes to events that have already taken place.
○ Counterfactual simulations allow one to simulate alternative or hypothetical scenarios which
were not present in the data.
3

• ALaSCA (Adaptable Large Scale Causal Analysis) has been
developed in Python, and leverages open source libraries for
core causal functions.
• The secret sauce in ALaSCA is its ability to work with
complex omics datasets modelling biological mechanisms.
4
ALaSCA* runs causal analysis on preclinical data
Causal Que
Literature
Study Data
etc.
Understanding
of Prior
Knowledge
ALaSCA Actionable
Output
Input Form
Highlighted genes from
traditional AI/ML
Limited Access Beta Program
Phase 1 – Proof of Concept on partner data and biological
mechanisms
Phase 2 – Deployment of ALaSCA for partner to use directly
Phase 3 – Integration of ALaSCA into partner’s analytics workflow
*Patent pending

Case Study: Type 1 Diabetes
5
Causal diagram of Type 1 Diabetes disease mechanism
Biological representation of disease mechanism Computational representation of disease mechanism
We utilised data from a Type 1 Diabetes case study to demonstrate ALaSCA’s causal inference and
counterfactual simulation capabilities:
● The proteomes of 10 healthy controls and 11 T1D patients were measured across 9 time points, from
birth to development of overt T1D (Liu, et al., 2018). The patients were selected based on the
presence of T1D susceptible HLA (human leukocyte antigen)-DR/DQ alleles through genotyping at
birth, therefore the cause of the disease was known.
● The following disease mechanism was translate into a causal diagram:

ALaSCA results:
Counterfactual simulation for hypothetical interventions
6
● Using ALaSCA’s counterfactual simulation capability,
we simulated a 50% increase in the antioxidant
protein SOD1 and compared the antibody levels
between the simulated T1D disease model with SOD1
activation (150% abundance) (blue) and observed
antibody levels from literature (black).
● The figure shows that an increase in abundance of the
antioxidant protein SOD1 can be seen to have a
protective (decreasing) effect on disease severity. This
is the same qualitative trend seen in literature.