SYNERGETIC

TIMELINE

Business Plan

TECHNOLOGY PLATFORM

STRATEGIC ROADMAP - 3 YEARS

GO-TO-MARKET: Senescence & Aging

ROADMAP: STRATEGIC GOALS & INITIATIVES

Calendar

Page 1

Explore

TECHNOLOGY PLATFORM

⁠

Add Topic

⁠

Topic Voting

Topic Voting

Done

Idea

Author

Vote

Upvotes

Upvoters

👍

👍

⁰

There are no rows in this table

⁠

TECHNOLOGY PLATFORM

Our platform builds upon three elements described in detail in the next section :

I. A central Software Application to plan, execute and monitor experiments on our models

II. In-Vitro models

III. In-Silico models

The technologies we leverage are :

Stem Cell (iPSC) Engineering

Instrumentation / Biology monitoring technologies

Deep Learning analytics & in-silico modelling

⁠

Fig 1. The Synergetic platform in a simplified form : Biology to Computer Model intervention-learning loop

Our technological platform is based on a closed-loop test-compare-learn approach.

Companies such as Zymergen, Gingko Bioworks, Recursion and recently Insitro applied similar integrated approaches with impressive levels of success.

We believe there is much more to be done before biology at large can benefit:

Better integration of computation and bioengineering technologies

Improved deep learning algorithms

Real efforts & progress towards interpretable AI

Improving the predictivity of stem cell engineering.

More importantly than any of the above points, our aim is to allow academic researchers, biotechs, and big pharma to participate in a large ecosystem based on these principles: in other words, we want to provide such a platform at scale.

I. Software Application

The central element of Synergetic offering is a software application allowing biologists to:

Plan and execute experiments on our in-vitro / in-silico models

Collect, visualize, and interpret the generated data

Collaborate worldwide with shared standardized data and protocols.

Factors of success in that space have been identified through interviews and competition analysis:

Simplicity and easy of use: users are largely non-experts and use few software applications

High Clarity / Readability of the predictions: quality visuals are a high incentive for use

Interface with academic publications: accessing knowledge from publications is currently a hurdle

Collaborative / Open Science features: repeating experiments already performed by other labs hundreds of times is a major pain point in scientific research.

In the age of web technology, social media, and collaboration, this is seen as highly inefficient by a large proportion of researchers.

Features :

Experiment planning & design support with experimental logical trees suggesting hypothesis of interest

Menu offering a choice of disease models: cell types, tissues, organoids ( sourced from our database of models)

Integrated in-vitro and in-silico digital twin model for predictive biology/in-silico experiments

Query system: keywords or complex query (options to filter genes proteins, compartment, function ex: mitochondria, lysosomes, RE, Golgi nucleus)

Previously Validated and Invalidated hypotheses (failed results) suggestions (with supporting data and analysis)

Experiments expected outcomes contrasted to real results

Data generated optionally feeding a Pooled repository from user community accessible to all (RNA-seq, proteomics, imaging

Databases dataset sourcing (RNA-seq, DNA-seq, proteomics, microscopy)

II. Biology - In-Vitro Models

A. Disease models

Synergetic builds model from induced pluripotent Stem Cells.

Induced pluripotent stem cells (iPS cells) are a type of pluripotent stem cells that can be derived from a various source of adult somatic cells by genetic reprogramming using a cocktail of genes identified as Yamanaka factors ( Oct4,Sox2,Klf4 and c-Myc (OSKM)).

In addition, two other genes have been discovered by Jean-Marc Lemaitre's team in Montpellier, Nanog and Lin28 that can be used in complement to the OSKM genes.

Once reprogrammed these news cells have the morphology and growth features of Embryonic cells (ES) by expressing embryonic genes.

Different methods exist to reprogram cells into iPS cells with different transforming success rates.

Recently a popular and safe method has been to use the non-integrative RNA Sendai virus, which allows us to over-express the reprogramming genes without the risk of random integration into the host genome that could lead to potential genomic issues.

What are the advantages of iPS cells over embryonic stem cells?

The main advantage of iPS cells is to allow:

The generation of stem cells without using human embryos. Research on human embryos is a major ethical concern and is very regulated.

The use of CRISP/Cas9 gene editing technologies in order to manipulate genetically cells to model diseases and differentiate a lineage with the DNA modifications.

Generate disease-specific models derived from patients with the genetic specificities maintained.

These cells offer a higher fidelity model and are a very valuable tool to study pathology and the physiology of various diseases in vitro to find treatments with higher safety and efficacy.

Are iPS cells similar to ES cells?

Yes, in many ways.

In culture they have similar morphology, proliferation, expression of pluripotency markers, long telomeres, can form teratoma, generate embryoid bodies.

In addition, they express stem cell surface markers and embryonic genes such as Nanog, Oct4, Sox2, TRA-1-60, TRA-1-81, SSEA-3, and SSEA-4, and have the capability to differentiate into various lineages.

How are iPS cells grown in culture?

In order to be able to produce iPS cells at the highest quality, protocols that are used must be standardized to offer reproducibility and similar controlled conditions between different laboratories. Lately, iPS cells are a favored tool to develop 3D cell culture organoids models which are closed to many organ systems in vitro.

Where can you obtain human iPS cells?

Sources for iPS cells are various depending on each laboratory and access to clinical samples from patients.

However, a reliable source is the European bank of induced pluripotent stem cells (EbiSC) which has a collection of high-quality human iPS cells available for research and disease modeling using a standardized procedure. This includes aged-matched and sex controls as well as pathological diseases (neurological, cardiovascular, macular, and many others).

Genetic engineering

While iPS cells offer a great deal to model disease a major barrier remains: the discrimination of the genetic background of the cells versus the effect of a causative mutation.

Luckily, great improvements have been made to facilitate genome editing techniques.

Tools such as CRISPR/Cas9 use a single non-sequence specific protein combined with a guiding RNA molecule that allows precise gene modification with high efficiency and accuracy.

This molecular tool enables the generation of controls and mutated cell lines to study the pathology caused by specific mutations including cardiac, neurodegenerative, and immunological diseases.

In addition, TAGS can also be added to increase spacialisation, timing and inducibility, using optogenetics.

Cell models

In order to built our complex models at a multi cellular level , we first have to create a simple but robust cell model.

This first model will be the first step to develop a proof of concept. We are starting with the most common cells used in laboratories in order to allow most academic labs to be able to participate. We then generate in-silico models of cells from the multi-omics and microscopy data.

A parallel approach involves starting from stem cells rather than adult cells, to recreate specific trajectories (differentiation into a specific lineage ex : fibroblasts, muscle cells, neurons ...ect). The process involves collecting data (multi omics and microscopy) along stages from a pluripotent stem cell state to a final differentiated state.

Organoids

Stem cells research over the years is increasing at a very high rate, offering the possibility of creating complex collections of multiple cells that conserve the genetic background of a patient's own tissues.

Organoids are small, self-organized 3D tissue cultures that are derived from stem cells or iPS cells that replicate the complexity of an organ including multiple cell types, structure, and function.

While many laboratories have discovered the right environment for the stem cells to derived into mini-organs (about the size of and hair up to five millimeters) such as brain, liver, kidney, intestine, and stomach, this is still an ongoing research area where a lot of improvements to be made.

Organoids have the ability to closely mimic human disease biology, which suggests a very high potential to revolutionize the field of drug discovery. We aim to improve these models beyond what they can currently provide through research and development.

Organoids models improvement pathway for a larger adoption principally depends on two research fronts that we aim to adress:

Organoids typically lack vasculature and immune cells, which means that:

They are limited in how large they can become without inducing cell death.

They cannot be yet used to study processes that require these components.

The cells in hPSC-derived organoids are relatively immature.

Multi-Organoids

The next development stage implies combining multiple organ systems together to study systemic interactions.

A high-fidelity multi-organ modular system provides a range of applications from clinical trials support and acceleration to personalized medicine.

B. Instrumentation

Microscopy / Imaging

Using recent advances in microscopy on biological models will allow us to visualize specific human cell fates transformation.

This approach allows to:

Track cells and to study how stem cells develop from early progenitors to final mature cells in multiple human organoid models.

Assess the impact of exchanging a single or multiples growth factor or of ECM (extracellular matrix) modifications on cellular behavior.

Determine how physical forces and cell shapes influence tissue differentiation or organ shape ( as new synthetic matrix components are developed )

Compatible microscopy approaches of interest are optical brightfield, live cell , confocal, lattice sheet microscope (3D).

After quality microscopy acquisition we will apply AI and machine learning technology to study morphological changes between states (healthy VS pathological) and increase the model's predictive power.

Multi-Omics

Biological systems rely on the axis DNA – RNA – protein information transfer paradigm that determines the function of any given organism.

Genomics, transcriptomics, and proteomics are now standard approaches in biology and opened the door to a range of discoveries.

In recent years:

Epigenomics and metabolomics have been added to the -omics arsenal to answer specific questions related to organisms functions.

Several of these techniques are now applicable at the single cell level including sc-RNA-seq.

Given the year on year advances in “omics” technologies, the volume and precision levels of information that can be generated in studies are expanding drastically.

We base our monitoring on these technologies and generate complex state maps of cells and tissues based on a multi-omics approaches.

Organs-on -chips

A natural next step from organoids models is to incorporate these models on chip systems with microfluidics, vasculature systems, gas exchanges, and flow.

This solution brings several advantages among which :

Standardization: models can be tested through a classical product quality testing process before being shipped. This addresses a key concern in biology research: reproducibility.

Advanced monitoring: instrumentation systems from microscopy to sequencing can be automatized on a single platform.

Modularity: plugging different organ types together becomes a possibility, this paves the way to the development of modular human biology "on-chips" with interesting applications in pre-clinical and clinical studies.

The connectivity is the most important part of this. What we envision is a process to producing pre-organ tissues that can develop naturally into high-fidelity organoids of different types. We are maximizing our capacity to make multiple organs much like our body does."

Improving human organ models is part of our core mission, these models can lead to faster, more predictive pre-clinical, and clinical research. Representative models can also reduce the need for animal testing, radically accelerate the time needed to put innovative therapies on the market, and bring readily applicable solutions to boost the development of personalized, precision medicine.

To build these models, we will:

Start with a single cell population (i.e. fibroblasts) inducing senescence in different ways

2. Model the fate of cells trough time (DNA, RNAseq, proteins, secreted proteins, and microscopy ).

3. Diversify the cell types

4. Start co-cultures of multi-cell-types tissues

5. Culture organoids (with increasing percentages of senescent cells) to be used as disease models.

End users can benefit from these high-fidelity disease models for a number of inquiries :

Study the factors leading a cell to undergo entering a senescent state instead of apoptosis

Study the % of senescent cells needed and average timeframes to disrupt tissue function

Test in-silico approches to eliminate senescent cells specifically (cell type specific)

Characterize key steps and states in the development of the disease

Connect the senescent organ with immune system models to monitor outcomes, such as recruitment and senescent cells clearance

Design a AI to study the differences of the healthy state VS the senescent state and predict possible therapeutic approches.

II. Deep Learning - In-Silico models

The core models we use are designed to build computer representations of cells and tissues.

They are designed to simulate biological processes and suggest the biological perturbations needed to reach specific states.

Modelling relies on a specifically designed pipeline of deep learning and programming.

⁠

A. A pipeline designed to reduce distance between cell states

Deep learning applied to multi-omics expression data is still in its infancy, but the future is bright.

Many previously untestable hypotheses can now be interrogated as deep learning enables analysis of increasing amounts of data generated by new technologies.

For example, the effects of cellular heterogeneity on basic biology and disease etiology can now be explored by single-cell RNA-seq.

Given a set of observed cell types in control (i.e real cells) and simulation, we aim to :

Define the distance between different cells by segmenting over omics and visual expression

Predict the perturbation response of specific cells by training a model that learns to generalize the response of the cells in the training set

Reduce the distance between unwanted cell states and desired cell states

Cell states are defined by multi-omics and microscopy depending on the type of modelling.

B. Deep Learning Models Pipeline

Our central aim with in-silico models is to identify and transform cell & tissue states.

To achieve this goal we are developing a pipeline based on 3 stages :

Graph-based modeling of known and inferred biological interactions

Deep unsupervised learning to identify cell states and sub-states

Adversarial models to suggest predictive states transformation

1. Graph-based modeling of known and inferred biological interactions

We are developing a graph-based pipeline to provide the initial structure for in-silico modeling, this pipeline consists of three stages :

A. Ontology generation : NLP methods

B. Equation based modeling

C. Deep Learning based approaches

Improving Equation based models

Intervention outcome prediction: efficacy, toxicity, cell faith (growth, apoptosis, differentiation, senescence)

2. Computer vision models to detect internal cells features

We are developing a deep learning pipeline to identify the internal features of cells, label them automatically, and represent them on screen.

This pipeline has 3 objectives :

A. Learn a latent space of visual features to be fed to the Cell Model

B. Perform automating segmentation

C. Simulate cells & tissues visually

3. Deep unsupervised learning to identify cell states and sub-states

Deep unsupervised learning models are well adapted to the task of defining meaningful states from large amounts of data.

The models we use also allow for latent space analysis to uncover interpretable classification rules, which offers a number of advantages compared to black box models.

The objective is to :

Learn a representation in the form of a latent space

Use vector arithmetics on the model’s latent space to identify meaningful dimensions

⁠

An illustrated example based on a VAE model .

4. Adversarial models to suggest predictive states transformations

A first approach for state transformation is based on the adversarial features of GANs.

The model consist of a generator and a discriminator neural networks that are trained jointly.

The generator and discriminator are trained concurrently so that the first learns to simulate realistic data and the second learns to differentiate between different types of data patterns (see figure below).

Initially the generator is fed noise and aims to generate realistic sequencing data samples, and thereby tries to deceive the discriminator into mistakenly classifying synthetic samples as real.

The discriminator classifies whether a given data point was drawn from the real data or whether it was synthetically generated. The weights of both networks are updated through back-propagation depending on the loss function.

⁠

This method is used to generate high fidelity computer representations of the cells we study. We reach that point once the discriminator is not able to differentiate between data coming from the real cells and the model ( and receptively predicts a 50% chance for both types.

Adversarial models can be adapted to identify biological perturbations that are likely to minimize the distance between :

Original tissues & biologically engineered tissues, the goal being to improve the representativity of cells, organoids and organs-on-chips

Disease cells & Healthy cells ( and tissues ) to uncover treatment options and relevant pathways

Different cell ( and tissue ) states along specific trajectories, for exemple stem cells to differentiated cells

Below is an exemple of a generator being trained to generate a healthy state from a disease state under the constraints of previously learned biologically plausible rules.

⁠

This approach, among others, allows us to not only differentiate between different cell states, but to suggest biologically coherent modification of those states. Beyond health and diseases, this is directly applicable to study and simulate cell trajectories between a variety of types and subtypes of interest.

Want to print your doc?
This is not the way.

Try clicking the ⋯ next to your doc name or using a keyboard shortcut (

CtrlP

) instead.