Ultra-deep sequencing allows detection of extremely rare mutational events that may be indicative of disease presence or recurrence. DEEPGEN utilizes the most advanced techniques to ensure that data output is the highest possible quality. DEEPGEN’s NGS process is outlined below:

• Receive standard blood sample (10-20ml)
• Nucleic acid extraction using the Qiagen QIAsymphony and an innovative extraction technology based on industry leading chemistry that achieves high efficiencies from standard input volumes of blood
• Run DEEPGEN assay on Illumina NovaSeq 6000

‍The DEEPGEN panel is unique in that the assay design provides both industry leading technical sensitivity and specificity in a large pan-cancer detection custom designed panel. The assay is comprised of a target panel of 272 genes identified to be related to cancer - through a proprietary machine learning algorithm. DEEPGEN ensures industry leading analytical and clinical performance. A combination of both proprietary and consensus markers, the panel contains genes in the U.S. National Comprehensive Cancer Network (NCCN) Guidelines, lung cancer and colorectal cancer (CRC), and emerging cancer biomarkers. DEEPGEN achieves a true limit of detection of 1/1000 genome copies, better than any other available technology to date.
‍
Technologies Used: Illumina NovaSeq 6000, Qiagen QIAsymphony, DEEPGEN custom assay

Output: FastQFiles

Using ultra-deep sequencing and proprietary analyses, extremely low level mutations can be detected and separated from the background noise of random mutations. Each base is evaluated to determine true pathogenic mutations from those that are nonpathogenic or due to sequencing errors.
‍
Technologies Used: DEEPGEN core engine (Deep signal processing, noise reduction, error correction, machine learning and cloud enabled bioinformatics system)
‍
Output: Mutational Patterns

DEEPGEN AI disease recognition technology utilizes both classical machine learning, and deep learning layers that have the ability to recognize relationalpatterns between individual features that correlate with certain disease profiles. DEEPGEN covers 15 cancer types including breast, colorectal, lung, ovarian, prostate, and more. This machine learning and AI infrastructure is what ultimately determines disease profiles in clinical samples that can then be used for multiple applications ranging from drug development to more general healthcare solutions.
‍
Technologies Used: Deep learning networks, DEEPGEN disease data sets from clinical trials

Output: Disease patterns and detection diagnostics

DEEPGEN analytics allows the user to understand mutational patterns by gene summaries, gene profiles including onco and tumor suppressor gene descriptions for mutations that were contained in the sample, as well as detected mutational targets that are FDA approved for guiding treatment decisions. DEEPGEN analytics also provides quality control tools that allow sample quality variables such as cfDNA input levels, plasma volumes, sequencing depth and total DNA copy counts by targets and amplicons. Additionally, easy overviews over clinical data allows for correlation between genomic and laboratory findings with clinical and demographic variables such as age, gender, ethnicity as well as comorbidities and diagnoses.