Search for: All records

Abstract CRISPR-Cas mediated genome engineering has revolutionized functional genomics. However, understanding of DNA repair following Cas-mediated DNA cleavage remains incomplete. Using Cas12a ribonucleoprotein genome editing in the fungal pathogen, Magnaporthe oryzae , we detail non-canonical DNA repair outcomes from hundreds of transformants. Sanger and nanopore sequencing analysis reveals significant variation in DNA repair profiles, ranging from small INDELs to kilobase size deletions and insertions. Furthermore, we find the frequency of DNA repair outcomes varies between loci. The results are not specific to the Cas-nuclease or selection procedure. Through Ku80 deletion analysis, a key protein required for canonical non-homologous end joining, we demonstrate activity of an alternative end joining mechanism that creates larger DNA deletions, and uses longer microhomology compared to C-NHEJ. Together, our results suggest preferential DNA repair pathway activity in the genome that can create different mutation profiles following repair, which could create biased genome variation and impact genome engineering and genome evolution. 

DNA double-strand breaks require repair or risk corrupting the language of life. To ensure genome integrity and viability, multiple DNA double-strand break repair pathways function in eukaryotes. Two such repair pathways, canonical non-homologous end joining and homologous recombination, have been extensively studied, while other pathways such as microhomology-mediated end joint and single-strand annealing, once thought to serve as back-ups, now appear to play a fundamental role in DNA repair. Here, we review the molecular details and hierarchy of these four DNA repair pathways, and where possible, a comparison for what is known between animal and fungal models. We address the factors contributing to break repair pathway choice, and aim to explore our understanding and knowledge gaps regarding mechanisms and regulation in filamentous pathogens. We additionally discuss how DNA double-strand break repair pathways influence genome engineering results, including unexpected mutation outcomes. Finally, we review the concept of biased genome evolution in filamentous pathogens, and provide a model, termed Biased Variation, that links DNA double-strand break repair pathways with properties of genome evolution. Despite our extensive knowledge for this universal process, there remain many unanswered questions, for which the answers may improve genome engineering and our understanding of genome evolution.

 

Recent years have witnessed the increasing penetration of wireless charging base stations in the workplace and public areas, such as airports and cafeterias. Such an emerging wireless charging infrastructure has presented opportunities for new indoor localization and identification services for mobile users. In this paper, we present QID, the first system that can identify a Qi-compliant mobile device during wireless charging in real-time. QID extracts features from the clock oscillator and control scheme of the power receiver and employs light-weight algorithms to classify the device. QID adopts a 2-dimensional motion unit to emulate a variety of multi-coil designs of Qi, which allows for fine-grained device fingerprinting. Our results show that QID achieves high recognition accuracy. With the prevalence of public wireless charging stations, our results also have important implications for mobile user privacy. 

SARS-CoV-2 and HIV-1 are RNA viruses that have killed millions of people worldwide. Understanding the similarities and differences between these two infections is critical for understanding disease progression and for developing effective vaccines and therapies, particularly for 38 million HIV-1+ individuals who are vulnerable to SARS-CoV-2 co-infection. Here, we utilized single-cell transcriptomics to perform a systematic comparison of 94,442 PBMCs from 7 COVID-19 and 9 HIV-1+ patients in an integrated immune atlas, in which 27 different cell types were identified using an accurate consensus single-cell annotation method. While immune cells in both cohorts show shared inflammation and disrupted mitochondrial function, COVID-19 patients exhibit stronger humoral immunity, broader IFN-I signaling, elevated Rho GTPase and mTOR pathway activities, and downregulated mitophagy. Our results elucidate transcriptional signatures associated with COVID-19 and HIV-1 that may reveal insights into fundamental disease biology and potential therapeutic targets to treat these viral infections. 

Recent trends towards large machine learning models require both training and inference tasks to be distributed. Considering the huge cost of training these models, it is imperative to unlock optimizations in computation and communication to obtain best performance. However, the current logical separation between computation and communication kernels in machine learning frameworks misses optimization opportunities across this barrier. Breaking this abstraction can provide many optimizations to improve the performance of distributed workloads. However, manually applying these optimizations requires modifying the underlying computation and communication libraries for each scenario, which is both time consuming and error-prone. Therefore, we present CoCoNet, which contains (i) a domain specific language to express a distributed machine learning program in the form of computation and communication operations, (ii) a set of semantics preserving transformations to optimize the program, and (iii) a compiler to generate jointly optimized communication and computation GPU kernels. Providing both computation and communication as first class constructs allows users to work on a high-level abstraction and apply powerful optimizations, such as fusion or overlapping of communication and computation. CoCoNet enabled us to optimize data-, model- and pipeline-parallel workloads in large language models with only a few lines of code. Our experiments show that CoCoNet significantly outperforms state-of-the-art distributed machine learning implementations.