WELCOME
Our vision: Revolutionizing healthcare with AI/ML-driven
multi-omics insights for early cancer detection and beyond, expanding to chronic diseases, acute infections, and personalized medicine.
Our mission: Advancing precision healthcare through scalable, non-invasive AI/ML solutions for early disease detection and monitoring.
About Us
Our Story
_edited.jpg)
OmicXHealth is a pioneering pre-seed startup focused on revolutionizing cancer detection through cutting-edge AI/ML and bioinformatics. We are currently working on a generalizable predictive test for early liver cancer detection, utilizing machine learning models trained on ctDNA methylation patterns from well-cited public datasets.
The early detection of cancer is one of the most effective ways to save lives, as survival rates drop significantly when cancer is diagnosed in its later stages. Our motivation stems from the understanding that through a simple, non-invasive test, such as a blood test, we can radically improve survival outcomes. Thanks to recent advances in cancer genomics and AI/ML, this vision is no longer a pipe dream but is well within reach, and we are committed to making it a reality.
We are powered by a team of industry experts, including specialists in bioinformatics, cancer genomics, and AI/ML technology. Together, we are developing innovative solutions to bring non-invasive, early diagnostics to the forefront of healthcare.
As we grow, we are always on the lookout for passionate experts, advisors, investors, and healthcare partners who share our vision. If you’re interested in collaborating with us, please reach out to learn more about how you can be part of this transformative mission.
Our work: Innovating Cancer detection with AI/ML and Epigenetics
Our work so far
At OmicXHealth, we are pioneering a cutting-edge solution to detect cancer early using AI-powered analysis of CpG methylation patterns. Our current focus is on hepatocellular carcinoma (HCC), one of the deadliest forms of liver cancer, with plans to expand to other cancer types in the future.
How It Works
-
Data Acquisition: We use publicly available bisulfite-sequenced datasets, such as GSE93203, consisting of over 1,200 HCC patients and 970 healthy controls. These datasets provide rich information on DNA methylation patterns, crucial for detecting cancer early.
-
Data Preprocessing: Using advanced bioinformatics tools, we clean and preprocess raw sequencing data to generate methylation profiles of CpG sites. These profiles represent critical epigenetic markers that may indicate the presence of cancer.
-
AI-Driven Model Training: Our machine learning models, including Random Forest and XGBoost, are trained on the methylation data to detect early-stage liver cancer with high accuracy. The models leverage the differential methylation patterns between cancerous and healthy samples to make predictions.
-
Results: Our models have demonstrated around 90% accuracy in detecting HCC, with balanced precision and recall, ensuring a reliable method for early detection.
What Sets Us Apart
-
Non-Invasive Detection: Leveraging liquid biopsies, our solution detects cancer through a simple blood test, providing a non-invasive and accessible alternative to traditional tissue biopsies.
-
Scalable and Generalizable: While our initial focus is on HCC, our approach is designed to generalize across datasets, providing a scalable platform for detecting multiple cancer types through epigenetic markers.
-
AI/ML Integration: By using AI to analyze complex epigenetic data, we ensure that our detection method is not only sensitive to early-stage cancers but also scalable for real-world clinical use.
.png)
The Road Ahead: Clinical Workflow
Fast paced, Innovative and Reliable Delivery
.png)
We are continuously refining our models and integrating additional datasets to further improve generalizability and performance. By the end of 2024, we aim to have a generalized model ready for clinical trials, with the potential to expand this approach to detect other cancers.
Stay tuned for more updates as we move towards real-world application and clinical validation!
What is Epigenetics
Epigenetics refers to changes in gene expression that do not involve alterations to the underlying DNA sequence. One of the key mechanisms of epigenetics is DNA methylation, which plays a significant role in turning genes on or off.
In the context of cancer detection, abnormal methylation patterns are often one of the earliest signs of cancer development. For example, hypermethylation (excessive methylation) can silence tumor suppressor genes, allowing cancer cells to proliferate unchecked. Similarly, hypomethylation (reduced methylation) can activate oncogenes, driving cancer growth.
At OmicXHealth, we focus on analyzing CpG methylation patterns, which are specific sites in the genome where DNA methylation occurs. These patterns serve as reliable biomarkers for early detection of cancer, allowing us to develop non-invasive blood tests that analyze these epigenetic markers to identify the presence of cancer.