Tinka Therapeutics had developed a machine learning framework that uses age-stratified transcriptomics data to generate de novo molecules for age-associated chronic diseases. We in-silico analysis the ADME and toxicology properties for all generated molecules. We prioritize molecules with a high probability of acting on molecular targets that are associated with longevity. Using our approach, it is possible to get lead compounds in a much shorter timeframe. Our future step is to synthesize these molecules and test them to be able to extend lifespan in vivo and move them forward into clinical applications.

Team:
CEO Prof. Dr. Collin Y. Ewald
CTO Noah Weber
CSO Tinka Vidovic

Our team:
Vincenzo Domina
Mike P. Sinn
Hermann Hampel
Evgeny Godlevsky
Mo Mahmood

The actionable aging clock makes information about the aging process accessible. This pathway information can be used to enable groundbreaking research into new drugs, repurposing of existing drugs, and other interventions to extend the healthspan of individuals. This information can also be used by a healthcare practitioner to recommend lifestyle changes with the goal of healthspan extension. The analysis is per individual, enabling personalized recommendations and treatment. Our analysis can be done for research from samples of deceased and living individuals, or to find interventions for living individuals and extend their healthspan.
Our AI predicts the biological age of an individual and the reasons why the prediction was made. The reasons are in the form of a gene regulatory networks/pathways activity map, which is then interpreted by professionals to decide what the appropriate interventions are.
We train our AI based on a unique multi-tissue/body fluid collection of single-cell RNA-seq data, because every tissue and every cell type/state may age differently.
We assemble existing data from public sources and collect private data from the Biobanks and universities we will form a collaboration with. We will supplement the procured data with our own scRNA-seq, to make sure our product goals can be satisfied.
With a significant dataset at hand, we will refine our AI to make accurate predictions that we can validate experimentally, publish papers as proof of our effectiveness, and secure our first clients for a POC. Once our production pipeline is in place, we move forward with a focus on growth which leads to additional data collection that promotes increasing accuracy of predictions and effectiveness of interventions that increment the number of clients which in turn produce more data, increased accuracy, and so on in a circle of growth.

Our team:
Bruno dos Santos - Data Scientist
Amir Mizrahi - Team Leader, Data Scientist
Enrico Glaab - Assistant Professor and Principal Investigator
Olivia Debnath - Computational Oncology Researcher
Dr. Ana Baroni MD. PhD. - Longevity, Regenerative & Precision Medicine Consultant

SudoClock, the winner of the LongHack HEALES award, is being jointly developed by Alex Trapp and Kejun "Albert" Ying, both members of Prof. Vadim Gladyshev's lab at Harvard Medical School. Our group is generally interested in developing and validating frameworks that enable high-resolution biological age profiling from molecular data. The most common types of biological clocks, known as epigenetic clocks, leverage age-associated changes in DNA cytosine methylation to enable biological age profiling in various tissues.

While these tools are each incredibly impactful for the longevity community, particularly in the interventions profiling space, these clocks are currently fragmented and hard to use. Indeed, they require having significant bioinformatics skills, reading complex methods sections in scientific papers, and browsing through obscure supplementary tables.

Our fundamental goal with SudoClock was to build a comprehensive and integrative framework that enables individuals, particularly newcomers to the longevity space, to calculate biological age from their samples of interest with minimal computational knowledge.

Presently, SudoClock integrates 7 human and 5 mouse epigenetic clocks, and we have plans to increase the number of clocks, species, and molecular modalities. Our web server, accessible at http://gladyshevlab.org:3838/SudoClock/, currently supports upload of processed high-throughput sequencing data, as well as popular microarray ("chip") platforms. Users can easily upload data in a variety of formats, as well as metadata that provides information about samples (species, experimental groups, chronological age, etc.).

Of note, we provide interactive visualization options, where users can explore the interactions between their variables of interest and biological ages according to different clocks. We also include an intuitive tutorial for the tool, as well as an atlas of current clocks built into SudoClock for easy reference.

Fundamentally, we believe in simplifying access to essential tools in longevity science, particularly methods for assessing age and the effect of rejuvenation interventions. We introduce SudoClock as an interactive, easy-to-use, and integrative analysis platform for biological age computation.

Our team:
Alex Trapp
Kejun "Albert" Ying

Rather than design drugs to target a few well-known aging genes, we propose an innovative systems approach to screen in silico a large number of chemical compounds and existing drugs to target all hallmarks of aging.

Our team:
Zhengdong Zhang
Jhih Rong Lin
Quanwei Zhang
Reza Jabal
Joydeep Mitra
Nha H Nguyen

AccuBio team: Maryna Polyakova Mark Melnykowycz