A5090783302- Glioma Diagnosis and Treatment
- Cancer Genomics and Diagnostics
- Genomics and Phylogenetic Studies
- Network Packet Processing and Optimization
- Algorithms and Data Compression
- Advanced biosensing and bioanalysis techniques
- Parallel Computing and Optimization Techniques
- Single-cell and spatial transcriptomics
- RNA and protein synthesis mechanisms
- Software Testing and Debugging Techniques
- Nanopore and Nanochannel Transport Studies
- Radiation Effects in Electronics
- DNA and Biological Computing
- MicroRNA in disease regulation
- Interconnection Networks and Systems
- Machine Learning in Bioinformatics
- Advanced Battery Technologies Research
- Biosensors and Analytical Detection
- RNA modifications and cancer
- Soft Robotics and Applications
- Bacteriophages and microbial interactions
- Chromosomal and Genetic Variations
- Cellular Automata and Applications
- Neuroblastoma Research and Treatments
- Molecular Biology Techniques and Applications
University of Michigan
2018-2025
Michigan Medicine
2021-2025
Michigan United
2021
University of Virginia
2014-2019
St. John's University
1982
Bell (Canada)
1979
Abstract Background Diffuse Midline Glioma (DMG) with the H3K27M mutation is a lethal childhood brain cancer, patients rarely surviving 2 years from diagnosis. Methods We conducted multi-site Phase 1 trial of imipridone ONC201 for children H3K27M-mutant glioma (NCT03416530). Patients enrolled on Arm D (n = 24) underwent serial lumbar puncture cell-free tumor DNA (cf-tDNA) analysis and all arms at University Michigan plasma collection. performed digital droplet polymerase chain reaction...
High-performance automata-processing engines are traditionally evaluated using a limited set of regular expression-rulesets. While expression rulesets valid real-world examples use cases for automata processing, they represent small proportion all automata-based computing. With the recent availability architectures and software frameworks many new applications have been found to benefit from processing. These show wide variety characteristics that differ prior, popular regular-expression...
Abstract We present SquiggleNet, the first deep-learning model that can classify nanopore reads directly from their electrical signals. SquiggleNet operates faster than DNA passes through pore, allowing real-time classification and read ejection. Using 1 s of sequencing data, classifier achieves significantly higher accuracy base calling followed by sequence alignment. Our approach is also requires an order magnitude less memory alignment-based approaches. distinguished human bacterial with...
ReadUntil enables Oxford Nanopore Technology's (ONT) sequencers to selectively sequence reads of target species in real-time. This efficient microbial enrichment for applications such as abundance estimation and is particularly beneficial metagenomic samples with a very high fraction non-target (> 99% can be human reads). However, read-until requires fast accurate software filter that analyzes short prefix read determines if it belongs microbe interest (target) or not. The baseline Read...
Reliability for general purpose processing on the GPU (GPGPU) is becoming a weak link in construction of reliable supercomputer systems. Because hardware protection expensive to develop, requires dedicated on-chip resources, and not portable across different architectures, efficiency software solutions such as redundant multithreading (RMT) must be explored. This paper presents real-world design evaluation automatic RMT hardware. We first describe compiler pass that automatically converts...
Pediatric high-grade glioma (pHGG) diagnosis portends poor prognosis and therapeutic monitoring remains difficult. Tumors release cell-free tumor DNA (cf-tDNA) into cerebrospinal fluid (CSF), allowing for potential detection of tumor-associated mutations by CSF sampling. We hypothesized that direct, electronic analysis cf-tDNA with a handheld platform (Oxford Nanopore MinION) could quantify patient-specific variant allele fraction (VAF) improved speed limit compared established methods.We...
The MinION is a recent-to-market handheld nanopore sequencer. It can be used to determine the whole genome of target virus in biological sample. Its Read Until feature allows us skip sequencing majority non-target reads (DNA/RNA fragments), which constitutes more than 99% all typical However, it does not have any on-board computing, significantly limits its portability.
The analysis of cell-free tumor DNA (ctDNA) and proteins in the blood patients with cancer potentiates a new generation non-invasive diagnostic approaches. However, confident detection tumor-originating markers is challenging, especially context brain tumors, where these analytes plasma are extremely scarce. Here, we apply sensitive single-molecule technology to profile multiple histone modifications on individual nucleosomes from diffuse midline glioma (DMG). system reveals epigenetic...
Reliability for general purpose processing on the GPU (GPGPU) is becoming a weak link in construction of reliable supercomputer systems. Because hardware protection expensive to develop, requires dedicated on-chip resources, and not portable across different architectures, efficiency software solutions such as redundant multithreading (RMT) must be explored. This paper presents real-world design evaluation automatic RMT hardware. We first describe compiler pass that automatically converts...
Finite automata have proven their usefulness in high-profile domains ranging from network security to machine learning. While prior work focused on applicability for purely regular expression workloads such as antivirus and rulesets, recent research has shown that can optimize the performance algorithms other areas learning even particle physics. Unfortunately, emulation traditional CPU architectures is fundamentally slow further bottlenecked by memory. In this paper, we present REAPR:...
Micron's new Automata Processor (AP) architecture exploits the very high and natural level of parallelism found in DRAM technologies to achieve native-hardware implementation non-deterministic finite automata (NFAs). The use technology implement NTA states provides capacity therefore provide extraordinary for pattern recognition. In this paper, we give an overview AP's architecture, programming applications.
Read alignment is a time-consuming step in genome sequencing analysis. The most widely used software for read alignment, BWA-MEM, and the recently published faster version BWA-MEM2 are based on seed-and-extend paradigm alignment. seeding of major bottleneck contributing ~40% to overall execution time when aligning whole human reads from Platinum Genomes dataset. This because both BWA-MEM use compressed index structure called FMD-Index, which results high bandwidth requirements, primarily due...
Automata Processing is an important kernel for many application domains, and challenging to accelerate using general purpose, von Neumann computers. New research into accelerators automata processing motivated the creation of ANMLZoo, a standardized benchmark suite that reflects modern use-cases processing. While researchers have adopted ANMLZoo as de-facto measure improvements in processing, certain drawbacks emerged after years use. In this work, we first examine opportunities improvement...
Automata processing has seen a resurgence in importance due to its usefulness for pattern matching and mining of "big data." While large-scale automata is known bottleneck von Neumann processors unpredictable memory accesses, spatial architectures excel at processing. Spatial can implement graphs by wiring together states reconfigurable arrays, allowing parallel state computation, point-to-point transitions on-chip. However, suffer from output constraints (up 255x commercial systems!) the...
We present MNCaRT, a comprehensive software ecosystem for the study and use of automata processing across hardware platforms. Tool support includes manipulation automata, execution complex machines, high-speed NFAs DFAs, compilation regular expressions. provide engines to execute on CPUs (with VASim Intel Hyperscan), GPUs custom DFA NFA engines), FPGAs an HDL translator). also introduce MNRL, open-source, general-purpose extensible state machine representation language developed MNCaRT. The...
Micron's Automata Processor (AP) efficiently emulates non-deterministic finite automata and has been shown to provide large speedups over traditional von Neumann execution for massively parallel, rule-based, data-mining pattern matching applications. We demonstrate the AP's ability generate high-quality energy efficient pseudo-random behavior use in number generation or chip simulation. By recognizing that transition rules become probabilistic when input characters are randomized, AP is also...
Molecular markers are essential for cancer diagnosis, clinical trial enrollment, and some surgical decision making, motivating ultra-rapid, intraoperative variant detection. Sequencing-based detection is considered the gold standard approach, but typically takes hours to perform due time-consuming DNA extraction, targeted amplification, library preparation times. In this work, we present a proof-of-principle approach sub-1 hour using real-time sequencers. By modifying existing protocols,...
Brill tagging is a classic rule-based algorithm for part-of-speech (POS) that assigns tags, such as nouns, verbs, adjectives, etc., to input tokens. Due the intense memory requirements of rule matching, CPU implementations have been found be slow. We show Micron's Automata Processor (AP) - new computing architecture can perform massively parallel pattern matching greatly accelerate second stage via template matching. The 218 contextual rules are first converted into regular expressions...
We present a general automata processing framework on FPGAs, which generates an RTL kernel for together with AXI and PCIe based I/O circuitry. implement the both local nodes cloud platforms (Amazon AWS Nimbix) novel features. A full performance comparison of proposed is conducted against state-of-the-art engines CPUs, GPUs, Micron’s Automata Processor using ANMLZoo benchmark suite some real-world datasets. Results show that FPGAs enable extremely high-throughput compared to von Neumann...