Exploring the expanding universe of small RNAs

Proteins are essential to life, and understanding their structure can facilitate a mechanistic understanding of their function. Through an enormous experimental effort 1 – 4 , the structures of around 100,000 unique proteins have been determined 5 , but this represents a small fraction of the billions of known protein sequences 6 , 7 . Structural coverage is bottlenecked by the months to years of painstaking effort required to determine a single protein structure. Accurate computational approaches are needed to address this gap and to enable large-scale structural bioinformatics. Predicting the three-dimensional structure that a protein will adopt based solely on its amino acid sequence—the structure prediction component of the ‘protein folding problem’ 8 —has been an important open research problem for more than 50 years 9 . Despite recent progress 10 – 14 , existing methods fall far short of atomic accuracy, especially when no homologous structure is available. Here we provide the first computational method that can regularly predict protein structures with atomic accuracy even in cases in which no similar structure is known. We validated an entirely redesigned version of our neural network-based model, AlphaFold, in the challenging 14th Critical Assessment of protein Structure Prediction (CASP14) 15 , demonstrating accuracy competitive with experimental structures in a majority of cases and greatly outperforming other methods. Underpinning the latest version of AlphaFold is a novel machine learning approach that incorporates physical and biological knowledge about protein structure, leveraging multi-sequence alignments, into the design of the deep learning algorithm. AlphaFold predicts protein structures with an accuracy competitive with experimental structures in the majority of cases using a novel deep learning architecture.

Over the last decade, approximately 20-30 nucleotide RNA molecules have emerged as critical regulators in the expression and function of eukaryotic genomes. Two primary categories of these small RNAs--short interfering RNAs (siRNAs) and microRNAs (miRNAs)--act in both somatic and germline lineages in a broad range of eukaryotic species to regulate endogenous genes and to defend the genome from invasive nucleic acids. Recent advances have revealed unexpected diversity in their biogenesis pathways and the regulatory mechanisms that they access. Our understanding of siRNA- and miRNA-based regulation has direct implications for fundamental biology as well as disease etiology and treatment.

An extensive repertoire of modifications is known to underlie the versatile coding, structural and catalytic functions of RNA, but it remains largely uncharted territory. Although biochemical studies indicate that N(6)-methyladenosine (m(6)A) is the most prevalent internal modification in messenger RNA, an in-depth study of its distribution and functions has been impeded by a lack of robust analytical methods. Here we present the human and mouse m(6)A modification landscape in a transcriptome-wide manner, using a novel approach, m(6)A-seq, based on antibody-mediated capture and massively parallel sequencing. We identify over 12,000 m(6)A sites characterized by a typical consensus in the transcripts of more than 7,000 human genes. Sites preferentially appear in two distinct landmarks--around stop codons and within long internal exons--and are highly conserved between human and mouse. Although most sites are well preserved across normal and cancerous tissues and in response to various stimuli, a subset of stimulus-dependent, dynamically modulated sites is identified. Silencing the m(6)A methyltransferase significantly affects gene expression and alternative splicing patterns, resulting in modulation of the p53 (also known as TP53) signalling pathway and apoptosis. Our findings therefore suggest that RNA decoration by m(6)A has a fundamental role in regulation of gene expression.