A BIOINFORMATICS PIPELINE FOR VIROLOGISTS POWERED BY DEEP MACHINE LEARNING

Funded by Small Business Innovation Research (SBIR) awards from the National Science Foundation (NSF), the National Institutes of Health (NIH/NIAID), and the Virginia Innovation Partnership Corporation (VIPC).

Funded by Small Business Innovation Research (SBIR) awards from the National Science Foundation (NSF) and the National Institutes of Health (NIH/NIAID)
Contact us
OUR FEATURES

Developing novel software solutions for virologists

Discover the future of genomics with GAT/ML by GATACA, a pioneering software pipeline that merges deep machine learning, natural language models, and bioinformatics algorithms within a high-performance dataflow system. GAT/ML doesn’t merely detect known variants—it’s designed to discover emerging variants in real-time, offering predictive insights into evolving quasispecies and shedding light on unseen pathogenic trajectories. Currently honed for Hepatitis and HIV analytics, GAT/ML extends its capabilities to any disease or sample characterized by sequence heterogeneity, making it a versatile tool applicable to virology, metagenomics, oncology, and others. By identifying and analyzing mixed populations, this alignment-free, reference-free system is equipped to revolutionize research and healthcare by predicting the unpredictable. Experience the meeting point of deep learning and genomic complexity with GAT/ML, and revolutionize the way you approach viral populations and diseases.

Customer focused
Value oriented
Quality obsessed
People centric
OUR FEATURES

Take control of your virology research & gain more insight faster

Data Metrics

GAT/ML streamlines your workflow, integrating standard and virus-centric metrics for comprehensive data quality and sequencing performance evaluation.

Data Metrics

From coverage depth and uniformity to GC bias and read ambiguity rate, gain in-depth insights into your NGS runs. Our ML model metrics reveal optimal classification boundaries, flag biases, uncover latent features, and more. Benefit from a detailed metrics report.

Mutation Capture

GAT/ML's hybrid assembly and ML algorithms employ a unique approach to deliver accurate intra-host variation representation, even among 99% similar real sample variants.

Mutation Capture

Our bias-free models transcend bioinformatics, revealing hidden mutation features.

Dynamic Databases

GAT/ML's dynamic database architecture overcomes the challenges posed by traditional, static data systems and the issues associated with data silos.

Dynamic Databases

GATACA’s databases are virus- and task-specific yet interoperable, enabling versatile capabilities and associations among genotype, anti-viral drug resistance and response, tropism, haplotype, quasispecies & more.

FUNCTIONAL TRAIT ANALYSIS

GAT/ML's models are trained on its large patient-specific database to identify clinical genomic attributes, predict their function, and link phenotypes to sequence analysis.

FUNCTIONAL TRAIT ANALYSIS

GAT/ML offers simultaneous monitoring of viral and host traits on the same quasispecies population over time.

integration of workflows

The key to a solid data management system is integrating the workflow, such as attaching annotations to sequences, creating data linkages, etc.

integration of workflows

By integrating the results of multiple experiments through proper normalization and cross-compiling, GAT/ML can discern batch effects and meaningful offsets. It unearths hidden patterns and trends, leveraging machine learning and predictive modeling to provide deeper insights.
nih logo
nsf logo
nih logo
cit logo
The HBV and HIV Problem
  • HBV and HIV are global, silent epidemics; many people are unaware of their infection(s)
  • 296 million people live with chronic HBV and 38.4 million with HIV; co-infection with HBV and HIV is common and harder to treat
  • HBV and HIV treatments are NOT A CURE -> Require life-long care and disease management
  • HBV and HIV evolve continuously in infected hosts as a population of molecular variants (quasispecies)
  • Quasispecies evolution leads to a selection of drug-resistant variants and clinical relapse – increasing risks and costs
  • Reference-free, alignment-free, reading frame-aware haplotype and quasispecies reconstructions and predictions
  • Deep learning directly from sequences, coupled with bioinformatics in a hybrid, automized approach
  • Unprecedented reliability through extensive testing and validations
our solution
the solution

Specialized NGS Bioinformatics Algorithms / Pipelines

Assembly and categorization of mixed genetic populations from NGS data
Current applications for HBV;
new training underway for HIV
Testing goals encompass benchmarking, confirming the natural occurrence of GAT/ML outputs, and validating predictive capabilities.
Reference-free, alignment-free, reading frame-aware haplotype and quasispecies reconstructions and predictions
Deep learning directly from sequences, coupled with bioinformatics in a hybrid, automized approach
Unprecedented reliability through extensive testing and validations
learn more
What we do
Additional Offerings
While GAT/ML is our primary software solution and offering (please contact us), we also provide the following services:
Unique data management and analytic methods

Custom Designed Analysis Tools

Unique data management and analytic methods that resolve high genetic heterogeneity and allow longitudinal analyses to track mutations and viral evolution

Our team of skilled writers and field-matched

Editing / Writing Services

Our team of skilled writers and field-matched bioinformatics experts will advise on the strategy, structure, and creation of your scientific documents and presentation materials.

Our team has extensive experience working with

Study Design Consultation

Our team has extensive experience working with, educating, and advocating alongside priority populations identified in the Viral Hepatitis Action Plan. 

Testimonials
Curious about what people say about us & our services?
"This GAT/ML project represents a leap in innovation, utilizing state-of-the-art machine learning to address a complex problem. With the PI's unique qualifications and exceptional team, the technique they're developing for HBV is promising. Their robust approach, underpinned by necessary controls and ground truth experiments, has us eagerly awaiting their progress."
insight from:A Bioinformatics Leader
insight from:
"Based on its design and purpose, GAT/ML v1.0 holds the promise to transform our work. It anticipates an unmet need and has potential utility across research, pharmaceuticals, clinics, and public health sectors. As it moves into beta testing, we're excited about the possibilities."
Feedback from:An Industry Insider
Feedback from:
"The GAT/ML software platform's capability to handle both single nucleotide variants and structural variants, along with other genomic complexities observed in quasi-species, truly sets it apart. The potential it holds for our field is vast and exciting."
Observation from:A Genomic Research Leader
Observation from:
"We are impressed by GATACA’s Phase I award HBV data showing the ability of the GAT/ML subtyping method to distinguish among HBV subtypes and ultra-sensitivity to create haplotypes from sequences with >99% similarity. We are also encouraged by its demonstrated extension to HIV variation, distinguishing among HIV subtype-level variation including env sequences. GAT/ML’s alignment-free components trained to learn species-specific viral sequence predictive of gaps and frame-shifts from indels promise to eliminate the need for codon- and frame-aware alignments. "
Feedback from:A Leader in Diagnostic Sciences
Feedback from:
"Recent optimizations to GAT/ML have significantly enhanced its performance, particularly its sensitivity to detect low-frequency variants. This has fueled our enthusiasm to dedicate time and resources to this collaboration. There is a pressing need for ML-powered tools for quasispecies analysis from NGS data - a need that currently lacks an adequate bioinformatics solution. Your algorithm addresses some crucial gaps, particularly in identifying minor ambiguities in the population, a factor with potentially major clinical implications."
Insight from:A Biomedical Informatics Specialist
Insight from:
Our Mission

To develop virus- and disease-specific molecular genome software solutions for discovery scientists

We understand your data analysis challenges and have the investigational tools to help you compete for funding, publications, patents, and new projects. Our software streamlines the discovery process with data organization and management, and applications integrating genetic, experimental, and clinical data.

Get in touch with us

Let's discuss now
Get in touch with us
We respond within 48 hours

We answer all emails and requests as they come in. If you have any questions about GAT/ML or beta testing or would like to place an order, please click the link below to give us a call.

Contact Information