bioinnovationbeacon.com

Category: AI in Biotechnology

  • DNA tech offers both data storage and computing functions

    DNA tech offers both data storage and computing functions

    Researchers from North Carolina State University and Johns Hopkins University have demonstrated a technology capable of a suite of data storage and computing functions — repeatedly storing, retrieving, computing, erasing or rewriting data — that uses DNA rather than conventional electronics. Previous DNA data storage and computing technologies could complete some but not all of these tasks.

    “In conventional computing technologies, we take for granted that the ways data are stored and the way data are processed are compatible with each other,” says project leader Albert Keung, co-corresponding author of a paper on the work. “But in reality, data storage and data processing are done in separate parts of the computer, and modern computers are a network of complex technologies.” Keung is an associate professor of chemical and biomolecular engineering and a Goodnight Distinguished Scholar at NC State.

    “DNA computing has been grappling with the challenge of how to store, retrieve and compute when the data is being stored in the form of nucleic acids,” Keung says. “For electronic computing, the fact that all of a device’s components are compatible is one reason those technologies are attractive. But, to date, it’s been thought that while DNA data storage may be useful for long-term data storage, it would be difficult or impossible to develop a DNA technology that encompassed the full range of operations found in traditional electronic devices: storing and moving data; the ability to read, erase, rewrite, reload or compute specific data files; and doing all of these things in programmable and repeatable ways.

    “We’ve demonstrated that these DNA-based technologies are viable, because we’ve made one.”

    The new technology is made possible by recent techniques that have enabled the creation of soft polymer materials that have unique morphologies.

    “Specifically, we have created polymer structures that we call dendricolloids — they start at the microscale, but branch off from each other in a hierarchical way to create a network of nanoscale fibers,” says Orlin Velev, co-corresponding author and the S. Frank and Doris Culberson Distinguished Professor of Chemical and Biomolecular Engineering at NC State. “This morphology creates a structure with a high surface area, which allows us to deposit DNA among the nanofibrils without sacrificing the data density that makes DNA attractive for data storage in the first place.”

    “You could put a thousand laptops’ worth of data into DNA-based storage that’s the same size as a pencil eraser,” Keung says.

    “The ability to distinguish DNA information from the nanofibers it’s stored on allows us to perform many of the same functions you can do with electronic devices,” says Kevin Lin, first author of the paper and a former Ph.D. student at NC State. “We can copy DNA information directly from the material’s surface without harming the DNA. We can also erase targeted pieces of DNA and then rewrite to the same surface, like deleting and rewriting information stored on the hard drive. It essentially allows us to conduct the full range of DNA data storage and computing functions. In addition, we found that when we deposit DNA on the dendricolloid material, the material helps to preserve the DNA.”

    “You could say that Keung’s team is providing the equivalent of microcircuits, and the dendricolloidal material that my team creates provides the circuit board,” says Velev. “Our NC State collaborator Adriana San Miguel helped us incorporate the materials into microfluidic channels that direct the flow of nucleic acids and reagents, allowing us to move data and initiate computing commands. Winston Timp’s lab at Johns Hopkins contributed their expertise on nanopore sequencing, which helps us directly read the data in RNA after copying it from DNA on the material’s surface. And James Tuck’s lab — also here at NC State — has developed algorithms that allow us to convert data into nucleic acid sequences and vice versa while controlling for potential errors.”

    The researchers have demonstrated that the new data storage and computing technology — which they call a “primordial DNA store and compute engine” — is capable of solving simple sudoku and chess problems. And testing suggests that it could store data securely for thousands of years in commercially available spaces without degrading the information-storing DNA.

    “What’s more, the dendrocolloidal host material itself is relatively inexpensive and easy to fabricate,” Velev says.

    “There’s a lot of excitement about molecular data storage and computation, but there have been significant questions about how practical the field may be,” says Keung. “We looked back at the history of computing and how the creation of ENIAC inspired the field. We wanted to develop something that would inspire the field of molecular computing. And we hope what we’ve done here is a step in that direction.”

    The paper, “A Primordial DNA Store and Compute Engine,” will be published Aug. 22 in the journal Nature Nanotechnology. The paper was co-authored by Kevin Volkel and Andrew Clark, former Ph.D. students at NC State; Cyrus Cao and Rachel Polak, Ph.D. students at NC State; Adriana San Miguel, an associate professor of chemical and biomolecular engineering at NC State; James Tuck, a professor of electrical and computer engineering at NC State; Winston Timp, an associate professor of biomedical engineering at Johns Hopkins University; and Paul Hook, a postdoctoral researcher at Johns Hopkins.

    The work was done with support from the National Science Foundation under grants 2027655 and 1901324.

    Keung and Tuck are co-founders of DNAli Data Technologies, so have potential interest in translating and commercializing DNA-based information systems. Keung, Volkel, Tuck, and Lin are inventors on patent application WO 2020/096679, which has been licensed to DNAli Data Technologies and from which some of this work is derived.

  • Using AI to find the polymers of the future

    Using AI to find the polymers of the future

    Nylon, Teflon, Kevlar. These are just a few familiar polymers — large-molecule chemical compounds — that have changed the world. From Teflon-coated frying pans to 3D printing, polymers are vital to creating the systems that make the world function better.

    Finding the next groundbreaking polymer is always a challenge, but now Georgia Tech researchers are using artificial intelligence (AI) to shape and transform the future of the field. Rampi Ramprasad’s group develops and adapts AI algorithms to accelerate materials discovery.

    This summer, two papers published in the Nature family of journals highlight the significant advancements and success stories emerging from years of AI-driven polymer informatics research. The first, featured in Nature Reviews Materials, showcases recent breakthroughs in polymer design across critical and contemporary application domains: energy storage, filtration technologies, and recyclable plastics. The second, published in Nature Communications, focuses on the use of AI algorithms to discover a subclass of polymers for electrostatic energy storage, with the designed materials undergoing successful laboratory synthesis and testing.

    “In the early days of AI in materials science, propelled by the White House’s Materials Genome Initiative over a decade ago, research in this field was largely curiosity-driven,” said Ramprasad, a professor in the School of Materials Science and Engineering. “Only in recent years have we begun to see tangible, real-world success stories in AI-driven accelerated polymer discovery. These successes are now inspiring significant transformations in the industrial materials R&D landscape. That’s what makes this review so significant and timely.”

    AI Opportunities

    Ramprasad’s team has developed groundbreaking algorithms that can instantly predict polymer properties and formulations before they are physically created. The process begins by defining application-specific target property or performance criteria. Machine learning (ML) models train on existing material-property data to predict these desired outcomes. Additionally, the team can generate new polymers, whose properties are forecasted with ML models. The top candidates that meet the target property criteria are then selected for real-world validation through laboratory synthesis and testing. The results from these new experiments are integrated with the original data, further refining the predictive models in a continuous, iterative process.

    While AI can accelerate the discovery of new polymers, it also presents unique challenges. The accuracy of AI predictions depends on the availability of rich, diverse, extensive initial data sets, making quality data paramount. Additionally, designing algorithms capable of generating chemically realistic and synthesizable polymers is a complex task.

    The real challenge begins after the algorithms make their predictions: proving that the designed materials can be made in the lab and function as expected and then demonstrating their scalability beyond the lab for real-world use. Ramprasad’s group designs these materials, while their fabrication, processing, and testing are carried out by collaborators at various institutions, including Georgia Tech. Professor Ryan Lively from the School of Chemical and Biomolecular Engineering frequently collaborates with Ramprasad’s group and is a co-author of the paper published in Nature Reviews Materials.

    “In our day-to-day research, we extensively use the machine learning models Rampi’s team has developed,” Lively said. “These tools accelerate our work and allow us to rapidly explore new ideas. This embodies the promise of ML and AI because we can make model-guided decisions before we commit time and resources to explore the concepts in the laboratory.”

    Using AI, Ramprasad’s team and their collaborators have made significant advancements in diverse fields, including energy storage, filtration technologies, additive manufacturing, and recyclable materials.

    Polymer Progress

    One notable success, described in the Nature Communications paper, involves the design of new polymers for capacitors, which store electrostatic energy. These devices are vital components in electric and hybrid vehicles, among other applications. Ramprasad’s group worked with researchers from the University of Connecticut.

    Current capacitor polymers offer either high energy density or thermal stability, but not both. By leveraging AI tools, the researchers determined that insulating materials made from polynorbornene and polyimide polymers can simultaneously achieve high energy density and high thermal stability. The polymers can be further enhanced to function in demanding environments, such as aerospace applications, while maintaining environmental sustainability.

    “The new class of polymers with high energy density and high thermal stability is one of the most concrete examples of how AI can guide materials discovery,” said Ramprasad. “It is also the result of years of multidisciplinary collaborative work with Greg Sotzing and Yang Cao at the University of Connecticut and sustained sponsorship by the Office of Naval Research.”

    Industry Potential

    The potential for real-world translation of AI-assisted materials development is underscored by industry participation in the Nature Reviews Materials article. Co-authors of this paper also include scientists from Toyota Research Institute and General Electric. To further accelerate the adoption of AI-driven materials development in industry, Ramprasad co-founded Matmerize Inc., a software startup company recently spun out of Georgia Tech. Their cloud-based polymer informatics software is already being used by companies across various sectors, including energy, electronics, consumer products, chemical processing, and sustainable materials.

    “Matmerize has transformed our research into a robust, versatile, and industry-ready solution, enabling users to design materials virtually with enhanced efficiency and reduced cost,” Ramprasad said. “What began as a curiosity has gained significant momentum, and we are entering an exciting new era of materials by design.”

  • Researchers develop AI model that predicts the accuracy of protein–DNA binding

    Researchers develop AI model that predicts the accuracy of protein–DNA binding

    A new artificial intelligence model developed by USC researchers and published in Nature Methods can predict how different proteins may bind to DNA with accuracy across different types of protein, a technological advance that promises to reduce the time required to develop new drugs and other medical treatments.

    The tool, called Deep Predictor of Binding Specificity (DeepPBS), is a geometric deep learning model designed to predict protein-DNA binding specificity from protein-DNA complex structures. DeepPBS allows scientists and researchers to input the data structure of a protein-DNA complex into an online computational tool.

    “Structures of protein-DNA complexes contain proteins that are usually bound to a single DNA sequence. For understanding gene regulation, it is important to have access to the binding specificity of a protein to any DNA sequence or region of the genome,” said Remo Rohs, professor and founding chair in the department of Quantitative and Computational Biology at the USC Dornsife College of Letters, Arts and Sciences. “DeepPBS is an AI tool that replaces the need for high-throughput sequencing or structural biology experiments to reveal protein-DNA binding specificity.”

    AI analyzes, predicts protein–DNA structures

    DeepPBS employs a geometric deep learning model, a type of machine-learning approach that analyzes data using geometric structures. The AI tool was designed to capture the chemical properties and geometric contexts of protein-DNA to predict binding specificity.

    Using this data, DeepPBS produces spatial graphs that illustrate protein structure and the relationship between protein and DNA representations. DeepPBS can also predict binding specificity across various protein families, unlike many existing methods that are limited to one family of proteins.

    “It is important for researchers to have a method available that works universally for all proteins and is not restricted to a well-studied protein family. This approach allows us also to design new proteins,” Rohs said.

    Major advance in protein-structure prediction

    The field of protein-structure prediction has advanced rapidly since the advent of DeepMind’s AlphaFold, which can predict protein structure from sequence. These tools have led to an increase in structural data available to scientists and researchers for analysis. DeepPBS works in conjunction with structure prediction methods for predicting specificity for proteins without available experimental structures.

    Rohs said the applications of DeepPBS are numerous. This new research method may lead to accelerating the design of new drugs and treatments for specific mutations in cancer cells, as well as lead to new discoveries in synthetic biology and applications in RNA research.

    About the study: In addition to Rohs, other study authors include Raktim Mitra of USC; Jinsen Li of USC; Jared Sagendorf of University of California, San Francisco; Yibei Jiang of USC; Ari Cohen of USC; and Tsu-Pei Chiu of USC; as well as Cameron Glasscock of the University of Washington.

    This research was primarily supported by NIH grant R35GM130376.

  • Researchers outline promises, challenges of understanding AI for biological discovery

    Researchers outline promises, challenges of understanding AI for biological discovery

    Machine learning is a powerful tool in computational biology, enabling the analysis of a wide range of biomedical data such as genomic sequences and biological imaging. But when researchers use machine learning in computational biology, understanding model behavior remains crucial for uncovering the underlying biological mechanisms in health and disease.

    In a recent article in Nature Methods, researchers at Carnegie Mellon University’s School of Computer Science propose guidelines that outline pitfalls and opportunities for using interpretable machine learning methods to tackle computational biology problems. The Perspectives article, “Applying Interpretable Machine Learning in Computational Biology — Pitfalls, Recommendations and Opportunities for New Developments,” is featured in the journal’s August special issue on AI.

    “Interpretable machine learning has generated significant excitement as machine learning and artificial intelligence tools are being applied to increasingly important problems,” said Ameet Talwalkar, an associate professor in CMU’s Machine Learning Department (MLD). “As these models grow in complexity, there is great promise not only in developing highly predictive models but also in creating tools that help end users understand how and why these models make certain predictions. However, it is crucial to acknowledge that interpretable machine learning has yet to deliver turnkey solutions to this interpretability problem.”

    The paper is a collaboration between doctoral students Valerie Chen in MLD and Muyu (Wendy) Yang in the Ray and Stephanie Lane Computational Biology Department. Chen’s earlier work critiquing the interpretable machine learning community’s lack of grounding in downstream use cases inspired the article, and the idea was developed through discussions with Yang and Jian Ma, the Ray and Stephanie Lane Professor of Computational Biology.

    “Our collaboration began with a deep dive into computational biology papers to survey the application of interpretable machine learning methods,” Yang said. “We noticed that many applications used these methods in a somewhat ad hoc manner. Our goal with this paper was to provide guidelines for more robust and consistent use of interpretable machine learning methods in computational biology.”

    One major pitfall the paper addresses is the reliance on a single interpretable machine learning method. Instead, the researchers recommend using multiple interpretable machine learning methods with diverse sets of hyperparameters and comparing their results to obtain a more comprehensive understanding of the model behavior and its underlying interpretations.

    “While some machine learning models seem to work surprisingly well, we often do not fully understand why,” Ma said. “In scientific domains like biomedicine, understanding why models work is crucial for discovering fundamental biological mechanisms.”

    The paper also warns against cherry-picking results when evaluating interpretable machine learning methods, as this can lead to incomplete or biased interpretations of scientific findings.

    Chen emphasized that the guidelines may have broader implications for a wider audience of researchers interested in applying interpretable machine-learning methods to their work.

    “We hope that machine learning researchers developing new interpretable machine learning methods and tools — particularly those working on explaining large language models — will carefully consider the human-centric aspects of interpretable machine learning,” Chen said. “This includes understanding who their target user is and how the method will be used and evaluated.”

    While understanding model behavior remains crucially important for scientific discovery and a fundamentally unsolved machine learning problem, the authors hope these challenges spur further interdisciplinary collaborations to facilitate the broader use of AI for scientific impact.

  • Cracking the code of life: new AI model learns DNA’s hidden language

    Cracking the code of life: new AI model learns DNA’s hidden language

    DNA contains foundational information needed to sustain life. Understanding how this information is stored and organized has been one of the greatest scientific challenges of the last century. With GROVER, a new large language model trained on human DNA, researchers could now attempt to decode the complex information hidden in our genome. Developed by a team at the Biotechnology Center (BIOTEC) of Dresden University of Technology, GROVER treats human DNA as a text, learning its rules and context to draw functional information about the DNA sequences. This new tool, published in Nature Machine Intelligence, has the potential to transform genomics and accelerate personalized medicine.

    Since the discovery of the double helix, scientists have sought to understand the information encoded in DNA. 70 years later, it is clear that the information hidden in the DNA is multilayered. Only 1-2 % of the genome consists of genes, the sequences that code for proteins.

    “DNA has many functions beyond coding for proteins. Some sequences regulate genes, others serve structural purposes, most sequences serve multiple functions at once. Currently, we don’t understand the meaning of most of the DNA. When it comes to understanding the non-coding regions of the DNA, it seems that we have only started to scratch the surface. This is where AI and large language models can help,” says Dr. Anna Poetsch, research group leader at the BIOTEC.

    DNA as a Language

    Large language models, like GPT, have transformed our understanding of language. Trained exclusively on text, the large language models developed the ability to use the language in many contexts.

    “DNA is the code of life. Why not treat it like a language?” says Dr. Poetsch. The Poetsch team trained a large language model on a reference human genome. The resulting tool named GROVER, or “Genome Rules Obtained via Extracted Representations,” can be used to extract biological meaning from the DNA.

    “GROVER learned the rules of DNA. In terms of language, we are talking about grammar, syntax, and semantics. For DNA this means learning the rules governing the sequences, the order of the nucleotides and sequences, and the meaning of the sequences. Like GPT models learning human languages, GROVER has basically learned how to ‘speak’ DNA,” explains Dr. Melissa Sanabria, the researcher behind the project.

    The team showed that GROVER can not only accurately predict the following DNA sequences but can also be used to extract contextual information that has biological meaning, e.g., identify gene promoters or protein binding sites on DNA. GROVER also learns processes that are generally considered to be “epigenetic,” i.e., regulatory processes that happen on top of the DNA rather than being encoded.

    “It is fascinating that by training GROVER with only the DNA sequence, without any annotations of functions, we are actually able to extract information on biological function. To us, it shows that the function, including some of the epigenetic information, is also encoded in the sequence,” says Dr. Sanabria.

    The DNA Dictionary

    “DNA resembles language. It has four letters that build sequences and the sequences carry a meaning. However, unlike a language, DNA has no defined words,” says Dr. Poetsch. DNA consists of four letters (A, T, G, and C) and genes, but there are no predefined sequences of different lengths that combine to build genes or other meaningful sequences.

    To train GROVER, the team had to first create a DNA dictionary. They used a trick from compression algorithms. “This step is crucial and sets our DNA language model apart from the previous attempts,” says Dr. Poetsch.

    “We analyzed the whole genome and looked for combinations of letters that occur most often. We started with two letters and went over the DNA, again and again, to build it up to the most common multi-letter combinations. In this way, in about 600 cycles, we have fragmented the DNA into ‘words’ that let GROVER perform the best when it comes to predicting the next sequence,” explains Dr. Sanabria.

    The Promise of AI in Genomics

    GROVER promises to unlock the different layers of genetic code. DNA holds key information on what makes us human, our disease predispositions, and our responses to treatments.

    “We believe that understanding the rules of DNA through a language model is going to help us uncover the depths of biological meaning hidden in the DNA, advancing both genomics and personalized medicine,” says Dr. Poetsch.

  • Method prevents an AI model from being overconfident about wrong answers

    Method prevents an AI model from being overconfident about wrong answers

    People use large language models for a huge array of tasks, from translating an article to identifying financial fraud. However, despite the incredible capabilities and versatility of these models, they sometimes generate inaccurate responses.

    On top of that problem, the models can be overconfident about wrong answers or underconfident about correct ones, making it tough for a user to know when a model can be trusted.

    Researchers typically calibrate a machine-learning model to ensure its level of confidence lines up with its accuracy. A well-calibrated model should have less confidence about an incorrect prediction, and vice-versa. But because large language models (LLMs) can be applied to a seemingly endless collection of diverse tasks, traditional calibration methods are ineffective.

    Now, researchers from MIT and the MIT-IBM Watson AI Lab have introduced a calibration method tailored to large language models. Their method, called Thermometer, involves building a smaller, auxiliary model that runs on top of a large language model to calibrate it.

    Thermometer is more efficient than other approaches — requiring less power-hungry computation — while preserving the accuracy of the model and enabling it to produce better-calibrated responses on tasks it has not seen before.

    By enabling efficient calibration of an LLM for a variety of tasks, Thermometer could help users pinpoint situations where a model is overconfident about false predictions, ultimately preventing them from deploying that model in a situation where it may fail.

    “With Thermometer, we want to provide the user with a clear signal to tell them whether a model’s response is accurate or inaccurate, in a way that reflects the model’s uncertainty, so they know if that model is reliable,” says Maohao Shen, an electrical engineering and computer science (EECS) graduate student and lead author of a paper on Thermometer.

    Shen is joined on the paper by Gregory Wornell, the Sumitomo Professor of Engineering who leads the Signals, Information, and Algorithms Laboratory in the Research Laboratory for Electronics, and is a member of the MIT-IBM Watson AI Lab; senior author Soumya Ghosh, a research staff member in the MIT-IBM Watson AI Lab; as well as others at MIT and the MIT-IBM Watson AI Lab. The research was recently presented at the International Conference on Machine Learning.

    Universal calibration

    Since traditional machine-learning models are typically designed to perform a single task, calibrating them usually involves one task-specific method. On the other hand, since LLMs have the flexibility to perform many tasks, using a traditional method to calibrate that model for one task might hurt its performance on another task.

    Calibrating an LLM often involves sampling from the model multiple times to obtain different predictions and then aggregating these predictions to obtain better-calibrated confidence. However, because these models have billions of parameters, the computational costs of such approaches rapidly add up.

    “In a sense, large language models are universal because they can handle various tasks. So, we need a universal calibration method that can also handle many different tasks,” says Shen.

    With Thermometer, the researchers developed a versatile technique that leverages a classical calibration method called temperature scaling to efficiently calibrate an LLM for a new task.

    In this context, a “temperature” is a scaling parameter used to adjust a model’s confidence to be aligned with its prediction accuracy. Traditionally, one determines the right temperature using a labeled validation dataset of task-specific examples.

    Since LLMs are often applied to new tasks, labeled datasets can be nearly impossible to acquire. For instance, a user who wants to deploy an LLM to answer customer questions about a new product likely does not have a dataset containing such questions and answers.

    Instead of using a labeled dataset, the researchers train an auxiliary model that runs on top of an LLM to automatically predict the temperature needed to calibrate it for this new task.

    They use labeled datasets of a few representative tasks to train the Thermometer model, but then once it has been trained, it can generalize to new tasks in a similar category without the need for additional labeled data.

    A Thermometer model trained on a collection of multiple-choice question datasets, perhaps including one with algebra questions and one with medical questions, could be used to calibrate an LLM that will answer questions about geometry or biology, for instance.

    “The aspirational goal is for it to work on any task, but we are not quite there yet,” Ghosh says.

    The Thermometer model only needs to access a small part of the LLM’s inner workings to predict the right temperature that will calibrate its prediction for data points of a specific task.

    An efficient approach

    Importantly, the technique does not require multiple training runs and only slightly slows the LLM. Plus, since temperature scaling does not alter a model’s predictions, Thermometer preserves its accuracy.

    When they compared Thermometer to several baselines on multiple tasks, it consistently produced better-calibrated uncertainty measures while requiring much less computation.

    “As long as we train a Thermometer model on a sufficiently large number of tasks, it should be able to generalize well across any new task, just like a large language model, it is also a universal model,” Shen adds.

    The researchers also found that if they train a Thermometer model for a smaller LLM, it can be directly applied to calibrate a larger LLM within the same family.

    In the future, they want to adapt Thermometer for more complex text-generation tasks and apply the technique to even larger LLMs. The researchers also hope to quantify the diversity and number of labeled datasets one would need to train a Thermometer model so it can generalize to a new task.

    This research was funded, in part, by the MIT-IBM Watson AI Lab.

  • Precision oncology via artificial intelligence on cancer biopsies

    Precision oncology via artificial intelligence on cancer biopsies

    A new generation of artificial intelligence (AI) tools designed to allow rapid, low-cost detection of clinically actionable genomic alterations directly from tumor biopsy slides has been developed by a team led by engineers and medical researchers at University of California San Diego.

    A paper describing the new AI protocol for examining routine biopsies, called DeepHRD, was recently published in the Journal of Clinical Oncology.

    Senior author Ludmil Alexandrov, Ph.D., professor of bioengineering and professor of cellular and molecular medicine at UC San Diego, says the new method is designed to save weeks and thousands of dollars from clinical oncology treatment workflows for breast and ovarian cancers.

    The team says their work represents an enormous step forward in the global efforts to eliminate the delays and health inequalities that have confounded the promise of precision medicine for cancer patients. Big picture: To develop new AI tools that can complement or replace the expensive and time-consuming genomic testing required to determine the best first-line cancer treatment specific for each individual patient.

    “A cancer patient today can expect to wait crucial weeks after their initial tumor diagnosis for a standard genomic test, resulting in life-threatening delays in treatment,” said Alexandrov. “It is very concerning that high costs and time delays render lifesaving treatment protocols inaccessible for most patients, disproportionately impacting resource-constrained settings.”

    At UC San Diego, this work represents a collaboration spanning all across campus, including the Department of Cellular and Molecular Medicine in the UC San Diego School of Medicine, the Shu Chien-Gene Lay Department of Bioengineering at the UC San Diego Jacobs School of Engineering, Institute of Engineering in Medicine, Department of Medicine, and the UC San Diego Moores Cancer Center.

    It was the potential of precision oncology to tailor an individual patient’s treatment options that motivated the collaborators, said Erik Bergstrom, Ph.D., lead author of the study and a postdoctoral researcher in Alexandrov’s lab, which bridges bioengineering and medicine.

    “Unfortunately, high costs, tissue requirements and slow turnaround times have hindered the widespread use of precision oncology, leading to suboptimal — potentially detrimental — treatment for cancer patients,” Bergstrom said. “We wanted to see if we could develop a completely different approach to resolve this serious issue by designing AI to circumvent the need for genomic testing.”

    Bergstrom said the collaborators focused on leveraging the minimum amount of patient information that is available early in the diagnostic process. He explained that virtually every cancer patient undergoes a tumor biopsy, a tissue sample that is routinely processed and examined through a light microscope. The process was developed in the late 19th century and is still the standard backbone of early clinical oncology workflows today.

    “Our AI, applied directly to a traditional tissue slide, allows accurate, instantaneous detection of cancer genomic biomarkers,” Bergstrom said. He explained that the team focused on AI identification of a specific biomarker for homologous recombination deficiency (HRD), a condition in which a cancerous cell loses a specific DNA damage repair mechanism.

    Bergstrom pointed out that patients with ovarian or breast cancers harboring HRD generally respond well to platinum and PARP (poly-ADP ribose polymerase) therapies, two common forms of chemotherapy.

    “This AI approach saves the patient critical time,” Alexandrov added. “Oncologists can prescribe treatment immediately after initial tissue diagnosis. Remarkably, the AI test has a negligible failure rate, while current genomic tests have a failure rate of 20 to 30 percent, necessitating re-testing, or even invasive re-biopsy.”

    The study’s co-senior author Scott Lippman, M.D., UC San Diego distinguished professor of medicine, Center for Engineering and Cancer, and Moores Cancer Center member, said the new technology will remove barriers of time and money to allow immediate, universal access and equality to actionable genomic biomarker detection — required for precision therapy — for people with advanced cancers. The extraordinary aspect of this breakthrough AI, is that it will benefit highly-informed, -resourced populations, and remarkably, will close the severe disparities ‘gap’ in precision medicine, especially in resource-constrained, remote regions worldwide where testing is not yet extant.

    “The era of precision oncology took off in the late 90s, but recent U.S. studies show that the vast majority of cancer patients are not getting FDA-approved precision therapy,” Lippman said. “And the prime reason is because they’re not getting tested. As a clinical oncologist — and I’ve been doing this for nearly 40 years — there is no question that this approach is the future of precision oncology.”

    The AI technology behind DeepHRD is protected by provisional patents through UC San Diego, which have been licensed to io9, a company with strong involvement by Alexandrov, Bergstrom and Lippman, and the goal to move this AI platform into the clinical arena as quickly as possible to make precision therapy real for patients with cancer by getting them onto the precise treatments they need faster. The authors expect that the same technology could be applied to most other genomic biomarkers and many forms of cancer.

    Affiliations: The authors have the affiliations noted above: Ludmil B. Alexandrov has appointments at UC San Diego Moores Cancer Center, School of Medicine Department of Cellular and Molecular Medicine, Sanford Stem Cell Institute and the Department of Bioengineering at the Irwin & Joan Jacobs School of Engineering. Erik N. Bergstrom is affiliated with Moores Cancer Center and Department of Cellular and Molecular Medicine. Scott M. Lippman has appointments at Moores Cancer Center, Department of Medicine; and is a Member & Board of Advisors and Co-director of the Center for Engineering and Cancer, Institute of Engineering in Medicine. Other co-authors are Ammal Abbasi and Marcos Díaz-Gay, both of Moores Cancer Center and Department of Cellular and Molecular Medicine and the Jacobs School of Engineering Department of Bioengineering; Loïck Galland and Sylvain Ladoire, both of the Department of Medical Oncology and the Platform of Transfer in Biological Oncology Centre, Georges-François Leclerc Cancer Center and University of Burgundy-Franche, France.

    Funding: This work was funded by U.S. National Institutes of Health grants R01ES032547 and U01DE033345 to Ludmil B. Alexandrov, and P30 CA023100 to Scott M. Lippman, as well as by a Curebound Targeted grant and UC San Diego start-up funding to Alexandrov. The research in this study was also supported by UC San Diego Sanford Stem Cell Institute.

  • A tool for visualizing single-cell data

    A tool for visualizing single-cell data

    Modern cutting-edge research generates enormous amounts of data, presenting scientists with the challenge of visualizing and analyzing it. Researchers at the Helmholtz Institute for RNA-based Infection Research (HIRI) in Würzburg and the Technical University of Applied Sciences Würzburg-Schweinfurt (THWS) have developed a tool for visualizing large data sets. The sCIRCLE tool allows users to explore single-cell analysis data in an interactive and user-friendly way. Their results have been published in the journal NAR Genomics and Bioinformatics.

    Antibiotic resistance is rising worldwide and poses a major challenge to healthcare systems. Understanding the underlying processes of how and why some bacteria develop such defenses is critical. Differences in bacterial gene expression — in which the information in a gene is read — could provide clues. Bacterial single-cell RNA sequencing, or “scRNA-seq,” has become an essential tool for analyzing these variations. This technology provides a detailed picture of gene expression in a single cell at a specific point in time. However, the method generates huge amounts of data that are difficult for researchers to visualize and analyze. Accordingly, there is a great need for new approaches and tools to display and understand this information.

    Researchers at the Helmholtz Institute for RNA-based Infection Research (HIRI) in Würzburg, a site of the Helmholtz Centre for Infection Research (HZI) in Braunschweig in cooperation with the Julius-Maximilians-Universität Würzburg (JMU), and designers at the Technical University of Applied Sciences Würzburg-Schweinfurt (THWS) have now developed such a tool. Their desktop application, called sCIRCLE (single-Cell Interactive Real-time Computer visualization for Low-dimensional Exploration), enables interactive 3D visualization of scRNA-seq data. It allows users to view single cells, enriched with metadata, from different angles and in real-time. A variety of filters and settings can be used to explore which genes have been expressed in particular cells at specific points in time. This capability allows researchers to focus on specific cell populations or genes of interest. “sCIRCLE is also a valuable communication tool for collaboratively examining data sets or presenting results,” says Lars Barquist. Barquist, a computational biologist who initiated the study, leads a research group at the Helmholtz Institute Würzburg. Barquist is also a professor at the University of Toronto in Canada. Additionally, sCIRCLE is already compatible with virtual reality devices, marking a step towards immersive 3D data visualization and interaction.

    A future-oriented cross-disciplinary collaboration

    The user-friendliness of sCIRCLE is notable: “The interface is easy to use and intuitively designed. sCIRCLE is therefore particularly helpful for biologists without extensive knowledge of bioinformatics,” adds Barquist. The designers and bioinformaticians worked together to make the data as simple and visually appealing as possible: “Collaboration between the disciplines of bioinformatics and design is extremely important as our data sets continue to grow and we need new ways to make them understandable,” says Barquist.

    To realize the project, THWS Master’s student Maximilian Seeger spent some time in the HIRI research group. There, he had the opportunity to work directly with the scientists. “This allowed me to understand what they wanted to explore in their data and develop an interface that would make this possible,” reports Maximilian Seeger. In test runs with data collected by HIRI researchers from different groups, the scientists tried out the tool and provided feedback. The research team has published the results in the journal NAR Genomics and Bioinformatics.

    “The joint development of sCIRCLE shows the importance of interdisciplinary and cross-institutional cooperation. I am very pleased we successfully combined the expertise of two Würzburg-based institutions in this project and look forward to future collaborations,” says Erich Schöls, professor for interactive media at THWS, who supervised Max Seeger together with Lars Barquist. “This is hopefully just the first step towards developing intuitive, interactive tools for data analysis,” says Barquist optimistically. “We will build on this foundation to make our tools even more immersive and user-friendly.” While this application currently offers only basic virtual reality integration, the team plans to create additional easy-to-use virtual spaces for data exploration in the future.

  • Genetic diagnostics of ultra-rare diseases

    Genetic diagnostics of ultra-rare diseases

    The majority of rare diseases have a genetic cause. The underlying genetic alteration can be found more and more easily, for example by means of exome sequencing (ES), leading to a molecular genetic diagnosis. ES is an examination of all sections of our genetic material (DNA) that code for proteins. As part of a Germany-wide multicenter study, ES data was collected from 1,577 patients and systematically evaluated. This made it possible to diagnose a total of 499 patients, with 34 patients showing new, previously unknown genetic diseases. The study thus makes a significant contribution to the initial description of new diseases. In addition, software based on the use of artificial intelligence (AI) was used for the first time on a broad scale to support clinical diagnosis. The “GestaltMatcher” AI system can assist in the assessment of facial features with regard to the classification of congenital genetic syndromes. The results of the study, in which 16 university locations were involved, have now been published in the journal Nature Genetics.

    Ultra-rare diseases require both multidisciplinary clinical expertise and comprehensive genetic diagnostics for optimal care. The three-year TRANSLATE NAMSE innovation fund project began at the end of 2017 with the aim of improving the care of those affected by means of modern diagnostic concepts. Researchers from 16 university hospitals analyzed the ES data of 1,577 patients, including 1,309 children, who presented to rare disease centers as part of TRANSLATE NAMSE. The aim of the project was to find the cause of the disease in as many patients as possible using innovative examination methods. A genetic cause of the rare disease was identified in 499 patients, 425 of whom were children. In total, the researchers found changes in 370 different genes. “We are particularly proud of the discovery of 34 new molecular diseases, which is a great example of knowledge-generating patient care at university hospitals,” says Dr. Theresa Brunet, one of the lead authors from the Institute of Human Genetics at the Klinikum rechts der Isar of the Technical University of Munich.

    What happens next with the unsolved cases?

    “We will examine the affected patients for whom we have not yet been able to find a diagnosis as part of the model project Genome Sequencing, or MVGenomSeq for short,” says Dr. Tobias Haack, Deputy Director of the Institute of Medical Genetics and Applied Genomics at the University Hospital of Tübingen. The MVGenomSeq builds on the success of the TRANSLATE NAMSE project and enables the analysis of clinical genomes at university hospitals throughout Germany. Unsolved cases can also be investigated in follow-up studies using new examination methods, such as long-read sequencing, which allows much longer DNA fragments to be analyzed. “Long-read sequencing enables us to find genetic changes that are difficult to detect and we assume that we will be able to make further diagnoses using this method,” says Dr. Nadja Ehmke, Head of Genome Diagnostics at Charité’s Institute of Medical Genetics and Human Genetics and one of the last authors.

    As part of the TRANSLATE NAMSE project, standardized procedures for extended genetic diagnostics for suspected rare diseases were also established at the participating rare disease centers, based on interdisciplinary case conferences. These were incorporated into standard care after the project was completed. “The interdisciplinary case conferences play an important role for those affected. This enables a comprehensive clinical characterization, which is relevant for the phenotype-based evaluation of the genetic data. In addition, the detected variants can be discussed in an interdisciplinary context,” says Dr. Magdalena Danyel, one of the first authors, who works as a specialist at the Institute of Medical Genetics and Human Genetics and a fellow of the Clinician Scientist Program of the Berlin Institute of Health (BIH) at Charité — Universitätsmedizin.

    Rare genetic diseases can sometimes be recognized by the face

    The researchers also investigated whether the supplementary use of machine learning and artificial intelligence (AI) tools improves diagnostic effectiveness and efficiency. To this end, the “GestaltMatcher” software developed by researchers in Bonn, which uses computer-assisted facial analysis to support the person using it in the diagnosis of rare diseases, was tested on a broad scale for the first time. The study used the sequence and image data of 224 people who had also consented to the computer-assisted analysis of their facial images, and it was shown that the AI-supported technology provides a clinical benefit.

    The GestaltMatcher AI can recognize abnormalities in the face and assign them to specific diseases. An important question when assessing genetic data is: Does the phenotype match the genotype? The AI can provide support here. “GestaltMatcher is like an expert opinion that we can provide to any medical professional in a matter of seconds. Early diagnosis is essential for those affected by rare diseases and their families. Supportive use of the software by pediatricians could already be useful in the case of abnormalities during the U7 screening at 21 to 24 months or U7a at 34 to 36 months,” says corresponding author Prof. Peter Krawitz, Director of the Institute for Genomic Statistics and Bioinformatics (IGSB) at the University Hospital Bonn (UKB), where the GestaltMatcher AI is being developed. Prof. Krawitz is also a member of the Cluster of Excellence ImmunoSensation2 and in the Transdisciplinary Research Areas (TRA) “Modeling” and “Life & Health” at the University of Bonn. The software and app can be made available to all doctors through the non-profit organization Arbeitsgemeinschaft für Gen-Diagnostik e.V. (AGD).

    Participating institutions

    In addition to the University Hospital Bonn (UKB) and the University of Bonn, the Charité-Universitätsmedizin Berlin, Klinikum rechts der Isar of the Technical University of Munich (TUM), University Hospital Düsseldorf, Ruhr University Bochum, University Hospital Dresden, University Hospital Essen, University Hospital Halle, University Hospital Hamburg Eppendorf, University Hospital Heidelberg, University Hospital Schleswig Holstein, LMU Hospital Munich, University Hospital RWTH Aachen, University Hospital Leipzig, University Hospital Tübingen and Stellenbosch University, Cape Town, South Africa were also involved

  • Using AI to scrutinize, validate theories on animal evolution

    Using AI to scrutinize, validate theories on animal evolution

    By harnessing the power of machine learning, researchers have constructed a framework for analyzing what factors most significantly contribute to a species’ genetic diversity.

    The study, recently published in the journal Molecular Phylogenetics and Evolution, suggests that the genetic variation of two species, the Brazilian sibilator frog and the granular toad, both amphibians native to northeastern Brazil, were shaped by different processes.

    Results showed that the genetic variation in the sibilator frog was shaped mostly by population demographic events in response to habitat changes that occurred over the last 100,000 years. In contrast, genetic diversity in the granular toad was mostly shaped by contemporary landscape factors — toads that are relatively more isolated, either by geographic distance or inhospitable habitat, were more likely to be genetically different.

    While previous investigations have explored the effects of historical demographic and landscape factors on genetic diversity of these amphibians, they were conducted with separate sets of data for these factors, making it difficult to discern which was the most important. Now, researchers involved with this paper are the first to use artificial intelligence to consider how both processes shape genetic diversity equally, rather than making manual assumptions about which may have been more vital.

    “Prior to this work, we had to ask questions independently because you couldn’t investigate both influences in the same framework,” said Bryan Carstens, co-author of the study and a professor in evolution, ecology and organismal biology at The Ohio State University. “What AI allows us to do is to simulate processes that are both happening ecologically in the present and during deep-time evolutionary events and compare those findings to the actual data that we collect from these frogs.”

    Due to the sheer amount of data that’s become available to geneticists and other wildlife biologists over the past few decades, it can be challenging for researchers to identify specific factors that might be important in certain experiments, said Carstens. But by integrating large swaths of information into simulations that can account for those elements in a single analysis, it’s possible to get a much more complete chronicle of a species’ development.

    “It takes a long time to build and train our AI models, but we wanted one l that could capture the range of potential variation in the species’ histories in a way that was as faithful as we could be to what we knew about the biology of the system,” said Carstens.

    For example, while the species this study investigated dwell in the same region, there are many differences in their natural histories. Despite both their eggs and larvae being fully aquatic, the sibulator frog reproduces continuously throughout the wet season and in underground chambers, while the granular toad’s reproductive events happen explosively because they are dependent on heavy rainfall.

    Combined with their machine learning approach, the researchers’ simulation determined their model scenarios were 100% supported regarding historical explanations for the sibilator frog’s expansion, and over 99% supported for those of the granular toad.

    One of the reasons their model is so accurate is due to its ability to account for recent demographic events, including measuring how events like human development or habitat change may have affected animal genetic diversity over a long period of time.

    But even when using AI, researchers have to be careful to avoid deceptive patterns in their results, said Carstens.

    “No analysis that we do is going to capture every single factor that has been important to these species over millions of years,” he said. “So we have to allow for a range of possibilities without making it so broad that essentially any model would be able to fit the data.”

    That said, as technological strides allow researchers to answer niche ecological questions and test new hypotheses, their work is a precursor to creating an upgraded machine learning framework that could be applied to unique investigations of other species, said Carstens.

    “We’re likely to continue using different combinations of these AI tools in different ways to try to understand evolutionary history,” said Carstens. “And as we keep learning, the tools we’re using will change, and they’ll evolve to be even better.”

    Emanuel M. Fonseca, who earned his doctorate from Ohio State in 2022, was a co-author. The study was supported by the Ohio Supercomputer Center, the U.S. National Science Foundation and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior in Brazil.