Opportunities and obstacles for deep learning in biology and medicine

Journal of The Royal Society Interface, Journal Year: 2018, Volume and Issue: 15(141), P. 20170387 - 20170387

Published: April 1, 2018

Deep learning describes a class of machine algorithms that are capable combining raw inputs into layers intermediate features. These have recently shown impressive results across variety domains. Biology and medicine data-rich disciplines, but the data complex often ill-understood. Hence, deep techniques may be particularly well suited to solve problems these fields. We examine applications biomedical problems-patient classification, fundamental biological processes treatment patients-and discuss whether will able transform tasks or if sphere poses unique challenges. Following from an extensive literature review, we find has yet revolutionize biomedicine definitively resolve any most pressing challenges in field, promising advances been made on prior state art. Even though improvements over previous baselines modest general, recent progress indicates methods provide valuable means for speeding up aiding human investigation. Though linking specific neural network's prediction input features, understanding how users should interpret models make testable hypotheses about system under study remains open challenge. Furthermore, limited amount labelled training presents some domains, as do legal privacy constraints work with sensitive health records. Nonetheless, foresee enabling changes at both bench bedside potential several areas biology medicine.

Language: Английский

---

The nuclear receptor superfamily: A structural perspective

Protein Science, Journal Year: 2018, Volume and Issue: 27(11), P. 1876 - 1892

Published: Aug. 15, 2018

Abstract Nuclear receptors (NRs) are a family of transcription factors that regulate numerous physiological processes such as metabolism, reproduction, inflammation, well the circadian rhythm. NRs sense changes in lipid metabolite levels to drive differential gene expression, producing distinct physiologic effects. This is an allosteric process whereby binding cognate ligand and specific DNA sequences drives recruitment diverse transcriptional co‐regulators at chromatin ultimately transactivation or transrepression target genes. Dysregulation NR signaling leads various malignances, metabolic disorders, inflammatory disease. Given their important role physiology ability respond small lipophilic ligands, have emerged valuable therapeutic targets. Here, we summarize discuss recent progress on understanding complex mechanism action NRs, primarily from structural perspective. Finally, suggest future studies improve our better design drugs by integrating multiple biophysical approaches.

Language: Английский

---

An improved MS2 system for accurate reporting of the mRNA life cycle

Nature Methods, Journal Year: 2017, Volume and Issue: 15(1), P. 81 - 89

Published: Nov. 13, 2017

Language: Английский

---

Intracellular mRNA transport and localized translation

Nature Reviews Molecular Cell Biology, Journal Year: 2021, Volume and Issue: 22(7), P. 483 - 504

Published: April 9, 2021

Language: Английский

---

Dual inhibition of MDMX and MDM2 as a therapeutic strategy in leukemia

Science Translational Medicine, Journal Year: 2018, Volume and Issue: 10(436)

Published: April 11, 2018

The tumor suppressor p53 is often inactivated via its interaction with endogenous inhibitors mouse double minute 4 homolog (MDM4 or MDMX) 2 (MDM2), which are frequently overexpressed in patients acute myeloid leukemia (AML) and other cancers. Pharmacological disruption of both these interactions has long been sought after as an attractive strategy to fully restore p53-dependent activity cancers wild-type p53. Selective targeting this pathway thus far limited MDM2-only small-molecule inhibitors, lack affinity for MDMX. We demonstrate that dual MDMX/MDM2 inhibition a stapled α-helical peptide (ALRN-6924), recently entered phase I clinical testing, produces marked antileukemic effects. ALRN-6924 robustly activates transcription at the single-cell single-molecule levels exhibits biochemical molecular biological on-target cells vitro vivo. Dual by inhibits cellular proliferation inducing cell cycle arrest apoptosis lines primary AML patient cells, including leukemic stem cell-enriched populations, disrupts functional clonogenic serial replating capacity. Furthermore, markedly improves survival xenograft models. Our study provides mechanistic insight support further testing therapeutic approach

Language: Английский

---

Recent advances in capillary electrophoresis‐mass spectrometry: Instrumentation, methodology and applications

Electrophoresis, Journal Year: 2018, Volume and Issue: 40(1), P. 79 - 112

Published: Sept. 27, 2018

Abstract Capillary electrophoresis (CE) offers fast and high‐resolution separation of charged analytes from small injection volumes. Coupled to mass spectrometry (MS), it represents a powerful analytical technique providing (exact) information enables molecular characterization based on fragmentation. Although hyphenation CE MS is not straightforward, much emphasis has been placed enabling efficient ionization user‐friendly coupling. Though several interfaces are now commercially available, research more robust interfacing with nano‐electrospray (ESI), matrix‐assisted laser desorption/ionization (MALDI) inductively coupled plasma (ICP) continues considerable results. At the same time, CE‐MS used in many fields, predominantly for analysis proteins, peptides metabolites. This review belongs series regularly published articles, summarizing 248 articles covering time between June 2016 May 2018. Latest developments as well instrumental such two‐dimensional systems detection mentioned. Furthermore, applications various CE‐modes including capillary zone (CZE), nonaqueous (NACE), gel (CGE) isoelectric focusing (CIEF) biological, pharmaceutical environmental summarized.

Language: Английский

---

Enhancer Histone Acetylation Modulates Transcriptional Bursting Dynamics of Neuronal Activity-Inducible Genes

Cell Reports, Journal Year: 2019, Volume and Issue: 26(5), P. 1174 - 1188.e5

Published: Jan. 1, 2019

Neuronal activity-inducible gene transcription correlates with rapid and transient increases in histone acetylation at promoters enhancers of activity-regulated genes. Exactly how modulates these genes has remained unknown. We used single-cell situ transcriptional analysis to show that Fos Npas4 are transcribed stochastic bursts mouse neurons membrane depolarization mRNA expression by increasing burst frequency. then expressed dCas9-p300 or dCas9-HDAC8 fusion proteins mimic block activity-induced locally enhancers. Adding increased prolonging duration resulted higher protein levels an elevation resting potential. Inhibiting reduced reducing frequency impaired experience-dependent induction the hippocampus vivo. Thus, tunes dynamics experience-regulated affect selective changes neuronal cellular function.

Language: Английский

---

Protein translation: biological processes and therapeutic strategies for human diseases

Signal Transduction and Targeted Therapy, Journal Year: 2024, Volume and Issue: 9(1)

Published: Feb. 23, 2024

Abstract Protein translation is a tightly regulated cellular process that essential for gene expression and protein synthesis. The deregulation of this increasingly recognized as critical factor in the pathogenesis various human diseases. In review, we discuss how deregulated can lead to aberrant synthesis, altered functions, disease progression. We explore key mechanisms contributing translation, including functional alterations factors, tRNA, mRNA, ribosome function. Deregulated leads abnormal expression, disrupted signaling, perturbed functions- all which contribute pathogenesis. development profiling techniques along with mass spectrometry-based proteomics, mRNA sequencing single-cell approaches have opened new avenues detecting diseases related errors. Importantly, highlight recent advances therapies targeting translation-related disorders their potential applications neurodegenerative diseases, cancer, infectious cardiovascular Moreover, growing interest lies targeted aimed at restoring precise control over diseased cells discussed. conclusion, comprehensive review underscores role its therapeutic target. Advancements understanding molecular deregulation, coupled therapies, offer promising improving outcomes Additionally, it will unlock doors possibility precision medicine by offering personalized deeper underpinnings future.

Language: Английский

---

Understanding RNA modifications: the promises and technological bottlenecks of the ‘epitranscriptome’

Open Biology, Journal Year: 2017, Volume and Issue: 7(5), P. 170077 - 170077

Published: May 1, 2017

The discovery of mechanisms that alter genetic information via RNA editing or introducing covalent modifications points towards a complexity in gene expression challenges long-standing concepts. Understanding the biology represents one next frontiers molecular biology. To this date, over 130 different have been identified, and improved mass spectrometry approaches are still adding to list. However, only recently has it possible map selected at single-nucleotide resolution, which created number exciting hypotheses about biological function modifications, culminating proposition ‘epitranscriptome’. Here, we review some technological advances rapidly developing field, identify conceptual discuss needed rigorously test specific modifications.

Language: Английский

---

Opportunities and obstacles for deep learning in biology and medicine

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2017, Volume and Issue: unknown

Published: May 28, 2017

Abstract Deep learning, which describes a class of machine learning algorithms, has recently showed impressive results across variety domains. Biology and medicine are data rich, but the complex often ill-understood. Problems this nature may be particularly well-suited to deep techniques. We examine applications biomedical problems—patient classification, fundamental biological processes, treatment patients—and discuss whether will transform these tasks or if sphere poses unique challenges. find that yet revolutionize definitively resolve any problems, promising advances have been made on prior state art. Even when improvement over previous baseline modest, we seen signs methods speed aid human investigation. More work is needed address concerns related interpretability how best model each problem. Furthermore, limited amount labeled for training presents problems in some domains, as do legal privacy constraints with sensitive health records. Nonetheless, foresee powering changes at both bench bedside potential several areas biology medicine.

Language: Английский

---