In the last few years, the deep learning (DL) computing paradigm has been deemed the Gold Standard in the machine learning (ML) community. Moreover, it has gradually become the most widely used computational approach in the field of ML, thus achieving outstanding results on several complex cognitive tasks, matching or even beating those provided by human performance. One of the benefits of DL is the ability to learn massive amounts of data. The DL field has grown fast in the last few years and it has been extensively used to successfully address a wide range of traditional applications. More importantly, DL has outperformed well-known ML techniques in many domains, e.g., cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, among many others. Despite it has been contributed several works reviewing the State-of-the-Art on DL, all of them only tackled one aspect of the DL, which leads to an overall lack of knowledge about it. Therefore, in this contribution, we propose using a more holistic approach in order to provide a more suitable starting point from which to develop a full understanding of DL. Specifically, this review attempts to provide a more comprehensive survey of the most important aspects of DL and including those enhancements recently added to the field. In particular, this paper outlines the importance of DL, presents the types of DL techniques and networks. It then presents convolutional neural networks (CNNs) which the most utilized DL network type and describes the development of CNNs architectures together with their main features, e.g., starting with the AlexNet network and closing with the High-Resolution network (HR.Net). Finally, we further present the challenges and suggested solutions to help researchers understand the existing research gaps. It is followed by a list of the major DL applications. Computational tools including FPGA, GPU, and CPU are summarized along with a description of their influence on DL. The paper ends with the evolution matrix, benchmark datasets, and summary and conclusion.

Sister chromatid cohesion is normally established in S phase in a process that depends on the cohesion establishment factor Eco1, a conserved acetyltransferase. However, due to the lack of known in vivo substrates, how Eco1 regulates cohesion is not understood. Here we report that yeast Eco1 and its human ortholog, ESCO1, both acetylate Smc3, a component of the cohesin complex that physically holds the sister chromatid together, at two conserved lysine residues. Mutating these lysine residues to a nonacetylatable form leads to increased loss of sister chromatid cohesion and genome instability in both yeast and human. In addition, we clarified that the acetyltransferase activity of Eco1 is essential for its function. Our study thus identified a molecular target for the acetyltransferase Eco1 and revealed that Smc3 acetylation is a conserved mechanism in regulating sister chromatid cohesion.

Although there has been considerable debate about whether paternal mitochondrial DNA (mtDNA) transmission may coexist with maternal transmission of mtDNA, it is generally believed that mitochondria and mtDNA are exclusively maternally inherited in humans. Here, we identified three unrelated multigeneration families with a high level of mtDNA heteroplasmy (ranging from 24 to 76%) in a total of 17 individuals. Heteroplasmy of mtDNA was independently examined by high-depth whole mtDNA sequencing analysis in our research laboratory and in two Clinical Laboratory Improvement Amendments and College of American Pathologists-accredited laboratories using multiple approaches. A comprehensive exploration of mtDNA segregation in these families shows biparental mtDNA transmission with an autosomal dominantlike inheritance mode. Our results suggest that, although the central dogma of maternal inheritance of mtDNA remains valid, there are some exceptional cases where paternal mtDNA could be passed to the offspring. Elucidating the molecular mechanism for this unusual mode of inheritance will provide new insights into how mtDNA is passed on from parent to offspring and may even lead to the development of new avenues for the therapeutic treatment for pathogenic mtDNA transmission.

ATM (ataxia telangiectasia-mutated) and ATR (ATM-Rad3-related) are proximal checkpoint kinases that regulate DNA damage response (DDR). Identification and characterization of ATM/ATR substrates hold the keys for the understanding of DDR. Few techniques are available to identify protein kinase substrates. Here, we screened for potential ATM/ATR substrates using phospho-specific antibodies against known ATM/ ATR substrates. We identified proteins cross-reacting to phospho-specific antibodies in response to DNA damage by mass spectrometry. We validated a subset of the candidate substrates to be phosphorylated in an ATM/ATR-dependent manner in vivo. Combining with a functional checkpoint screen, we identified proteins that belong to the ubiquitin-proteasome system (UPS) to be required in mammalian DNA damage checkpoint control, particularly the G 1 cell cycle checkpoint, thus revealing protein ubiquitylation as an important regulatory mechanism downstream of ATM/ATR activation for checkpoint control.

No abstract