Linear DNA Sequence Research Articles

Comparative analysis of predicted DNA secondary structures infers complex human centromere topology

Secondary structures are non-canonical arrangements of nucleic acids due to intra-strand interactions, including base pairing, stacking, or other higher-order features that deviate from the standard double-helical conformation. While these structures are extensively studied in RNA, they can also form when DNA becomes single stranded, creating topological roadblocks that can impact essential DNA-based processes such as replication, transcription, and repair, ultimately affecting genome stability. The availability of a complete linear sequence of human genomes, including repetitive loci, enables the prediction of DNA secondary structures comparing across various regions. Here, we evaluate the intrinsic properties of linear single-stranded DNA sequences derived from sampling specialized human loci such as centromeres, pericentromeres, ribosomal DNA (rDNA), and coding regions from the CHM13 genome. Our comparative analysis of predicted secondary structures across human chromosomes revealed the heightened presence, complexity, and instability of secondary structures within the centromere, which gradually decreased toward the pericentromere onto chromosomes' arms, on average lowest in coding regions. Notably, centromeric repeats exhibited the highest level of topological complexity within both the active and divergent domains, even when compared to other repetitive tandem satellites, such as rDNA in acrocentric chromosomes. Our findings provide evidence of the intrinsic self-hybridizing properties of centromere repeats, which are capable of generating complex topological structures that may functionally correlate with chromosome missegregation, especially when centromeric chromatin is disrupted. Processes such as long non-coding RNA transcription, recombination, and other mechanisms that dechromatinize and unwind stretches of linear DNA in these regions create invivo opportunities for the DNA acrobatics hereby predicted.

SEGUID v2: Extending SEGUID checksums for circular, linear, single- and double-stranded biological sequences.

Synthetic biology involves combining different DNA fragments, each containing functional biological parts, to address specific problems. Fundamental gene-function research often requires cloning and propagating DNA fragments, such as those from the iGEM Parts Registry or Addgene, typically distributed as circular plasmids.Addgene's repository alone offers around 150,000 plasmids. To ensure data integrity, cryptographic checksums can be calculated for the sequences. Each sequence has a unique checksum, making checksums useful for validation and quick lookups of associated annotations. For example, the SEGUID checksum, uniquely identifies protein sequences with a 27-character string. The original SEGUID, while effective for protein sequences and single-stranded DNA (ssDNA), is not suitable for circular and double-stranded DNA (dsDNA) due to topological differences. Challenges include how to uniquely represent linear dsDNA, circular ssDNA, and circular dsDNA. To meet these needs, we propose SEGUID v2, which extends the original SEGUID to handle additional types of sequences. SEGUID v2 produces orientation and rotation invariant checksums for single-stranded, double-stranded, possibly staggered, linear, and circular DNA and RNA sequences. Customizable alphabets allow for other types of sequences. In contrast to the original SEGUID, which uses Base64, SEGUID v2 uses Base64url to encode the SHA-1 hash. This ensures SEGUID v2 checksums can be used as-is in filenames, regardless of platform, and in URLs, with minimal friction. SEGUID v2 is readily available for major programming languages, distributed under the MIT license. JavaScript package seguid is available on npm, Python package seguid on PyPi, R package seguid on CRAN, and a Tcl script on GitHub. These tools, along with documentation, examples, and an online SEGUID Calculator , can be found at https://www.seguid.org .

Molecular evolution of the mammalian kinetochore complex.

Mammalian centromeres are satellite-rich chromatin domains that serve as sites for kinetochore complex assembly. Centromeres are highly variable in sequence and satellite organization across species, but the processes that govern the co-evolutionary dynamics between rapidly evolving centromeres and their associated kinetochore proteins remain poorly understood. Here, we pursue a course of phylogenetic analyses to investigate the molecular evolution of the complete kinetochore complex across primate and rodent species with divergent centromere repeat sequences and features. We show that many protein components of the core centromere associated network (CCAN) harbor signals of adaptive evolution, consistent with their intimate association with centromere satellite DNA and roles in the stability and recruitment of additional kinetochore proteins. Surprisingly, CCAN and outer kinetochore proteins exhibit comparable rates of adaptive divergence, suggesting that changes in centromere DNA can ripple across the kinetochore to drive adaptive protein evolution within distant domains of the complex. Our work further identifies kinetochore proteins subject to lineage-specific adaptive evolution, including rapidly evolving proteins in species with centromere satellites characterized by higher-order repeat structure and lacking CENP-B boxes. Thus, features of centromeric chromatin beyond the linear DNA sequence may drive selection on kinetochore proteins. Overall, our work spotlights adaptively evolving proteins with diverse centromere-associated functions, including centromere chromatin structure, kinetochore protein assembly, kinetochore-microtubule association, cohesion maintenance, and DNA damage response pathways. These adaptively evolving kinetochore protein candidates present compelling opportunities for future functional investigations exploring how their concerted changes with centromere DNA ensure the maintenance of genome stability.

Imputation of 3D genome structure by genetic–epigenetic interaction modeling in mice

Gene expression is known to be affected by interactions between local genetic variation and DNA accessibility, with the latter organized into three-dimensional chromatin structures. Analyses of these interactions have previously been limited, obscuring their regulatory context, and the extent to which they occur throughout the genome. Here, we undertake a genome-scale analysis of these interactions in a genetically diverse population to systematically identify global genetic–epigenetic interaction, and reveal constraints imposed by chromatin structure. We establish the extent and structure of genotype-by-epigenotype interaction using embryonic stem cells derived from Diversity Outbred mice. This mouse population segregates millions of variants from eight inbred founders, enabling precision genetic mapping with extensive genotypic and phenotypic diversity. With 176 samples profiled for genotype, gene expression, and open chromatin, we used regression modeling to infer genetic–epigenetic interactions on a genome-wide scale. Our results demonstrate that statistical interactions between genetic variants and chromatin accessibility are common throughout the genome. We found that these interactions occur within the local area of the affected gene, and that this locality corresponds to topologically associated domains (TADs). The likelihood of interaction was most strongly defined by the three-dimensional (3D) domain structure rather than linear DNA sequence. We show that stable 3D genome structure is an effective tool to guide searches for regulatory elements and, conversely, that regulatory elements in genetically diverse populations provide a means to infer 3D genome structure. We confirmed this finding with CTCF ChIP-seq that revealed strain-specific binding in the inbred founder mice. In stem cells, open chromatin participating in the most significant regression models demonstrated an enrichment for developmental genes and the TAD-forming CTCF-binding complex, providing an opportunity for statistical inference of shifting TAD boundaries operating during early development. These findings provide evidence that genetic and epigenetic factors operate within the context of 3D chromatin structure.

Imputation of 3D genome structure by genetic-epigenetic interaction modeling in mice.

Gene expression is known to be affected by interactions between local genetic variation and DNA accessibility, with the latter organized into three-dimensional chromatin structures. Analyses of these interactions have previously been limited, obscuring their regulatory context, and the extent to which they occur throughout the genome. Here, we undertake a genome-scale analysis of these interactions in a genetically diverse population to systematically identify global genetic-epigenetic interaction, and reveal constraints imposed by chromatin structure. We establish the extent and structure of genotype-by-epigenotype interaction using embryonic stem cells derived from Diversity Outbred mice. This mouse population segregates millions of variants from eight inbred founders, enabling precision genetic mapping with extensive genotypic and phenotypic diversity. With 176 samples profiled for genotype, gene expression, and open chromatin, we used regression modeling to infer genetic-epigenetic interactions on a genome-wide scale. Our results demonstrate that statistical interactions between genetic variants and chromatin accessibility are common throughout the genome. We found that these interactions occur within the local area of the affected gene, and that this locality corresponds to topologically associated domains (TADs). The likelihood of interaction was most strongly defined by the three-dimensional (3D) domain structure rather than linear DNA sequence. We show that stable 3D genome structure is an effective tool to guide searches for regulatory elements and, conversely, that regulatory elements in genetically diverse populations provide a means to infer 3D genome structure. We confirmed this finding with CTCF ChIP-seq that revealed strain-specific binding in the inbred founder mice. In stem cells, open chromatin participating in the most significant regression models demonstrated an enrichment for developmental genes and the TAD-forming CTCF-binding complex, providing an opportunity for statistical inference of shifting TAD boundaries operating during early development. These findings provide evidence that genetic and epigenetic factors operate within the context of 3D chromatin structure.

Chromatin structure and dynamics: one nucleosome at a time.

Eukaryotic genomes store information on many levels, including their linear DNA sequence, the posttranslational modifications of its constituents (epigenetic modifications), and its three-dimensional folding. Understanding how this information is stored and read requires multidisciplinary collaborations from many branches of science beyond biology, including physics, chemistry, and computer science. Concurrent recent developments in all these areas have enabled researchers to image the genome with unprecedented spatial and temporal resolution. In this review, we focus on what single-molecule imaging and tracking of individual proteins in live cells have taught us about chromatin structure and dynamics. Starting with the basics of single-molecule tracking (SMT), we describe some advantages over insitu imaging techniques and its current limitations. Next, we focus on single-nucleosome studies and what they have added to our current understanding of the relationship between chromatin dynamics and transcription. In celebration of Robert Feulgen's ground-breaking discovery that allowed us to start seeing the genome, we discuss current models of chromatin structure and future challenges ahead.

Morphine Re-arranges Chromatin Spatial Architecture of Primate Cortical Neurons

The expression of linear DNA sequence is precisely regulated by the three-dimensional (3D) architecture of chromatin. Morphine-induced aberrant gene networks of neurons have been extensively investigated; however, how morphine impacts the 3D genomic architecture of neurons is still unknown. Here, we applied digestion-ligation-only high-throughput chromosome conformation capture (DLO Hi-C) technology to investigate the effects of morphine on the 3D chromatin architecture of primate cortical neurons. After receiving continuous morphine administration for 90 days on rhesus monkeys, we discovered that morphine re-arranged chromosome territories, with a total of 391 segmented compartments being switched. Morphine altered over half of the detected topologically associated domains (TADs), most of which exhibited a variety of shifts, followed by separating and fusing types. Analysis of the looping events at kilobase-scale resolution revealed that morphine increased not only the number but also the length of differential loops. Moreover, all identified differentially expressed genes from the RNA sequencing data were mapped to the specific TAD boundaries or differential loops, and were further validated for changed expression. Collectively, an altered 3D genomic architecture of cortical neurons may regulate the gene networks associated with morphine effects. Our finding provides critical hubs connecting chromosome spatial organization and gene networks associated with the morphine effects in humans.

An all-deoxyribonucleic acid circuit for detection of human telomerase activity in solution and on paper

We report on the design of a simple all-DNA circuit with dual functions of signal amplification and signal reporting and its use for detection of human telomerase activity from cancer cells. The system utilizes a catalytic hairpin assembly (CHA) reaction for amplification, which produces split G-quadruplex outputs that assemble to form complete guanine quadruplex structures as reporting modules. As designed, a linear DNA sequence (the target) functions as a catalyst to drive cyclic programmed assembly of two hairpins, producing a DNA duplex with two guanine-rich sequences that assemble to form a complete Gq structure. The formation of the Gq element allows either fluorescence or colorimetric detection of the target. Examples are provided to demonstrate fluorescence detection of cancer cells’ telomerase activities in solution and the first example of a CHA-modulated colorimetric assay for detecting telomerase activities of cancer cells using a simple paper device.

Abstract 5797: Simultaneous profiling of genetic variants, 3D genome, GpC methyltransferase footprints, DNA methylation, and transcriptome in a single assay

Abstract Cis-regulatory elements (CREs) play instrumental roles in the regulation of mammalian gene expression. Sets of transcriptional factors bring together multiple CREs, which can be up to megabases away, to initiate and maintain the gene expression. How to measure the activity of CREs and their coordination with the genetic variants over large genomic distances to regulate gene expression is the central question for regulatory genomics, which is still limited by the current sequencing approaches. We report a method, TAGMe-seq, which simultaneously captures the single nucleotide polymorphisms (SNPs), chromosome conformation, DNA methylation, and GpC methyltransferase footprints from a single molecule, together with transcriptome in the same assay.To demonstrate the performance of TAGMe-seq, we applied it to the IMR-90 and GM12878 cell lines, both of which have been comprehensively profiled publically. TAGMe-seq showed high concordances with the state-of-art mono-omics assays across all the molecular measurements at the same cell types. We next asked if the genomic regions linearly separated but positioned proximity in space would have coordinated footprint status. We analyzed the GCH (H=A, C, or T) methyltransferase footprint on TAGMe-seq reads separated in chromatin loop anchors. Indeed, the GCH methyltransferase footprint in TAGMe-seq read pairs mapping to separated loop anchors showed a significantly higher correlation than that in the shuffled read pairs from the same genomic regions. To further demonstrate the utility of TAGMe-seq, we utilized it to analyze the long-range (>1kb) allele-specific GpC methyltransferase footprint. We separated TAGMe-seq reads into two groups by their parent-of-origin at the heterozygous SNPs (SNP anchors). Naturally, the other ends of the read pairs were also separated into two alleles, even though they are not overlapped with the SNPs and mapped over large genomic distances (non-SNP anchors). Thus, we are able to analyze the long-range allele-specific GCH methyltransferase footprint. We identified three groups of allele-specific footprint loci based on their differences of GCH accessibility between the two alleles at both anchors, or only at SNP anchors, or only at non-SNP anchors, which showed distinct enrichment levels of TF binding. Overall, combining multiple assays in TAGMe-seq allowed us to uncover the coordinated GpC methyltransferase footprint and long-range allele-specific footprint at genomic regions that are far away in linear DNA sequences but spatially close in the nucleus, which will eventually pave the road to dissect the mechanisms of gene-regulatory aberrations in complex diseases, such as cancer. Citation Format: Hailu Fu, Haizi Zheng, Louis J. Muglia, Li Wang, Yaping Liu. Simultaneous profiling of genetic variants, 3D genome, GpC methyltransferase footprints, DNA methylation, and transcriptome in a single assay [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5797.

Characterizing collaborative transcription regulation with a graph-based deep learning approach.

Human epigenome and transcription activities have been characterized by a number of sequence-based deep learning approaches which only utilize the DNA sequences. However, transcription factors interact with each other, and their collaborative regulatory activities go beyond the linear DNA sequence. Therefore leveraging the informative 3D chromatin organization to investigate the collaborations among transcription factors is critical. We developed ECHO, a graph-based neural network, to predict chromatin features and characterize the collaboration among them by incorporating 3D chromatin organization from 200-bp high-resolution Micro-C contact maps. ECHO predicted 2,583 chromatin features with significantly higher average AUROC and AUPR than the best sequence-based model. We observed that chromatin contacts of different distances affected different types of chromatin features' prediction in diverse ways, suggesting complex and divergent collaborative regulatory mechanisms. Moreover, ECHO was interpretable via gradient-based attribution methods. The attributions on chromatin contacts identify important contacts relevant to chromatin features. The attributions on DNA sequences identify TF binding motifs and TF collaborative binding. Furthermore, combining the attributions on contacts and sequences reveals important sequence patterns in the neighborhood which are relevant to a target sequence's chromatin feature prediction.

GRAFIMO: Variant and haplotype aware motif scanning on pangenome graphs.

Transcription factors (TFs) are proteins that promote or reduce the expression of genes by binding short genomic DNA sequences known as transcription factor binding sites (TFBS). While several tools have been developed to scan for potential occurrences of TFBS in linear DNA sequences or reference genomes, no tool exists to find them in pangenome variation graphs (VGs). VGs are sequence-labelled graphs that can efficiently encode collections of genomes and their variants in a single, compact data structure. Because VGs can losslessly compress large pangenomes, TFBS scanning in VGs can efficiently capture how genomic variation affects the potential binding landscape of TFs in a population of individuals. Here we present GRAFIMO (GRAph-based Finding of Individual Motif Occurrences), a command-line tool for the scanning of known TF DNA motifs represented as Position Weight Matrices (PWMs) in VGs. GRAFIMO extends the standard PWM scanning procedure by considering variations and alternative haplotypes encoded in a VG. Using GRAFIMO on a VG based on individuals from the 1000 Genomes project we recover several potential binding sites that are enhanced, weakened or missed when scanning only the reference genome, and which could constitute individual-specific binding events. GRAFIMO is available as an open-source tool, under the MIT license, at https://github.com/pinellolab/GRAFIMO and https://github.com/InfOmics/GRAFIMO.

Embracing Metagenomic Complexity with a Genome-Free Approach.

ABSTRACTA central paradigm in microbiome data analysis, which we term the genome-centric paradigm, is that a linear (non-branching) DNA sequence is the ideal representation of a microbial genome. This representation is natural, as microbes indeed have non-branching genomes. Tremendous discoveries in microbiology were made under this paradigm, but is it always optimal for microbiome research? In this Commentary, we claim that the realization of this paradigm in metagenomic assembly, a fundamental step in the “metagenomics analysis pipeline,” suboptimally models the extensive genomic variability present in the microbiome. We outline our efforts to address these issues with a “genome-free” approach that eschews linear genomic representations in favor of a pan-metagenomic graph.

Toward an understanding of the relation between gene regulation and 3D genome organization

BackgroundHigh‐order chromatin structure has been shown to play a vital role in gene regulation. Previously we identified two types of sequence domains, CGI (CpG island) forest and CGI prairie, which tend to spatially segregate, but to different extent in different tissues. Here we aim to further quantify the association of domain segregation with gene regulation and therefore differentiation.MethodsBy means of the published RNA‐seq and Hi‐C data, we identified tissue‐specific genes and quantitatively investigated how their regulation is relevant to chromatin structure. Besides, two types of gene networks were constructed and the association between gene pair co‐regulation and genome organization is discussed.ResultsWe show that compared to forests, tissue‐specific genes tend to be enriched in prairies. Highly specific genes also tend to cluster according to their functions in a relatively small number of prairies. Furthermore, tissue‐specific forest‐prairie contact formation was associated with the regulation of tissue‐specific genes, in particular those in the prairie domains, pointing to the important role of gene positioning, in the linear DNA sequence as well as in 3D chromatin structure, in gene regulatory network formation.ConclusionWe investigated how gene regulation is related to genome organization from the perspective of forest‐prairie spatial interactions. Since unlike compartments A and B, forest and prairie are identified solely based on sequence properties. Therefore, the simple and uniform framework (forest‐prairie domain segregation) provided here can be utilized to further understand the chromatin structure changes as well as the underlying biological significances in different stages, such as tumorgenesis.

Toolbox Accelerating Glycomics (TAG): Glycan Annotation from MALDI-TOF MS Spectra and Mapping Expression Variation to Biosynthetic Pathways.

Glycans present extraordinary structural diversity commensurate with their involvement in numerous fundamental cellular processes including growth, differentiation, and morphogenesis. Unlike linear DNA and protein sequences, glycans have heterogeneous structures that differ in composition, branching, linkage, and anomericity. These differences pose a challenge to developing useful software for glycomic analysis. To overcome this problem, we developed the novel Toolbox Accelerating Glycomics (TAG) program. TAG consists of three units: ‘TAG List’ creates a glycan list that is used for database searching in TAG Expression; ‘TAG Expression’ automatically annotates and quantifies glycan signals and draws graphs; and ‘TAG Pathway’ maps the obtained expression information to biosynthetic pathways. Herein, we discuss the concepts, outline the TAG process, and demonstrate its potential using glycomic expression profile data from Chinese hamster ovary (CHO) cells and mutants lacking a functional Npc1 gene (Npc1 knockout (KO) CHO cells). TAG not only drastically reduced the amount of time and labor needed for glycomic analysis but also detected and quantified more glycans than manual analysis. Although this study was limited to the analysis of N-glycans and free oligosaccharides, the glycomic platform will be expanded to facilitate the analysis of O-glycans and glycans of glycosphingolipids.

Multi-modal and multiscale imaging approaches reveal novel cardiovascular pathophysiology in Drosophila melanogaster.

ABSTRACTEstablishing connections between changes in linear DNA sequences and complex downstream mesoscopic pathology remains a major challenge in biology. Herein, we report a novel, multi-modal and multiscale imaging approach for comprehensive assessment of cardiovascular physiology in Drosophila melanogaster. We employed high-speed angiography, optical coherence tomography (OCT) and confocal microscopy to reveal functional and structural abnormalities in the hdp2 mutant, pre-pupal heart tube and aorta relative to controls. hdp2 harbor a mutation in wupA, which encodes an ortholog of human troponin I (TNNI3). TNNI3 variants frequently engender cardiomyopathy. We demonstrate that the hdp2 aortic and cardiac muscle walls are disrupted and that shorter sarcomeres are associated with smaller, stiffer aortas, which consequently result in increased flow and pulse wave velocities. The mutant hearts also displayed diastolic and latent systolic dysfunction. We conclude that hdp2 pre-pupal hearts are exposed to increased afterload due to aortic hypoplasia. This may in turn contribute to diastolic and subtle systolic dysfunction via vascular-heart tube interaction, which describes the effect of the arterial loading system on cardiac function. Ultimately, the cardiovascular pathophysiology caused by a point mutation in a sarcomeric protein demonstrates that complex and dynamic micro- and mesoscopic phenotypes can be mechanistically explained in a gene sequence- and molecular-specific manner.

Fast and Secure Medical Image Encryption Based on Non Linear 4D Logistic Map and DNA Sequences (NL4DLM_DNA).

This paper proposes an innovative image cryptosystem algorithm using the properties of the block encryption, 4D logistic map and DNA systems. Multiple key sequences are generated and pixel substitution is performed by using nonlinear 4D logistic map, then encryption is performed by using DNA rules to ensure that the different blocks are encrypted securely. The results of the experiment indicate that the proposed Non Linear 4D Logistic Map and DNA (NL4DLM_DNA) sequence based algorithm gives better performance, which is analyzed on the basis of security, quality, attack resilience, diffusion and running time as compared to some previous works.

Characterization of Structural Variations in the Context of 3D Chromatin Structure.

Chromosomes located in the nucleus form discrete units of genetic material composed of DNA and protein complexes. The genetic information is encoded in linear DNA sequences, but its interpretation requires an understanding of three-dimensional (3D) structure of the chromosome, in which distant DNA sequences can be juxtaposed by highly condensed chromatin packing in the space of nucleus to precisely control gene expression. Recent technological innovations in exploring higher-order chromatin structure have uncovered organizational principles of the 3D genome and its various biological implications. Very recently, it has been reported that large-scale genomic variations may disrupt higher-order chromatin organization and as a consequence, greatly contribute to disease-specific gene regulation for a range of human diseases. Here, we review recent developments in studying the effect of structural variation in gene regulation, and the detection and the interpretation of structural variations in the context of 3D chromatin structure.

365 Multivalent chromatin looping orchestrates cellular reprogramming for advanced gene therapy

The discovery of cellular plasticity, an epigenetic phenomenon by which differentiated cells alter their identity through reprogramming, dedifferentiation or transdifferentiation has led to a paradigm shift in regenerative medicine for treatment of genetic diseases and injuries. Current surgical methods pose high mortality rates making stem cell-based therapy an attractive option. HOX genes, a set of transcription factor genes, play key roles during stem cell renewal, multilineage differentiation and reprogramming of differentiated cells. Delineating the molecular mechanisms of HOX genes regulation in diverse cellular identities will provide clues to modulate cell fate for development of novel, more effective stem cell-based treatments. Recent studies highlight that a cell identity is determined not only by specific gene expression profile and posttranslational modifications but largely by its characteristic three-dimensional (3D) genomic organization. Chromatin maintains a high-order topology organizing genes into structural domains through long-range chromatin looping that is essential for controlled gene regulation. Existing techniques to study the link between chromatin looping and gene regulation require either large alterations of the linear DNA sequence or prior knowledge of loop-mediating factors. We have pioneered a technology to form multivalent chromosomal contacts and loops in 3D space at will. Previous studies revealed that primary adult fibroblasts retain several HOX gene expression patterns similar to that in embryonic cells. We will affect chromosomal remodeling using our new multiplexed chromatin looping system to examine how 3D architectural reorganization of chromatin at the Hox loci in human dermal fibroblasts can alter cellular identities (via reprogramming and transdifferentiation) with the potential for reproducible and efficient clinical-grade cell production. This study will provide insights for linking chromatin architecture to cutaneous gene expression that affects cellular plasticity in both adult and embryonic cells that may have transformative outcomes for regenerative medicine.

Three-dimensional chromatin packing and positioning of plant genomes.

Information and function of a genome are not only decorated with epigenetic marks in the linear DNA sequence but also in their non-random spatial organization in the nucleus. Recent research has revealed that three-dimensional (3D) chromatin organization is highly correlated with the functionality of the genome, contributing to many cellular processes. Driven by the improvements in chromatin conformation capture methods and visualization techniques, the past decade has been an exciting period for the study of plants' 3D genome structures, and our knowledge in this area has been substantially advanced. This Review describes our current understanding of plant chromatin organization and positioning beyond the nucleosomal level, and discusses future directions.

Abstract 2357: Utilizing biological and protein structure-guided features to improve driver mutation discovery

Abstract Distinguishing between driver and passenger somatic mutations to pinpoint genetic alterations leading to oncogenesis still presents significant challenges. To meet these challenges, computational tools have been developed as effective filters, pruning most of the somatic mutations to a shortlist of high-priority, functional candidates for experimental validation. Most tools include searching for genes or pathways having mutation rates higher than explained by chance, mutations in conserved regions, or genes with neighboring mutations on the linear DNA or protein sequence. Recently, there has been a shift to utilize tertiary/quaternary protein structures to identify mutations clustering proximal to each other in 3D space. Such enrichment of mutations can indicate specific domains critical to normal protein function and when mutated, can drive tumor initiation and progression. HotSpot3D, a protein structure-based tool, identifies clusters enriched with proximal mutations within proteins. Though HotSpot3D has been valuable in identifying clusters of residues that are important to cancer, it does not distinguish the driving potential or structural impact of different mutations within a cluster nor does it consider the physical impact of different amino acid substitutions at the same site. The prediction power of HotSpot3D in distinguishing driver mutations from passenger mutations can be improved if spatial clustering considers physical/biological features proximal to mutations in significant clusters as well as the specific amino acid substitutions of mutations. We have created a machine learning algorithm that further prioritizes putative driver mutations found in HotSpot3D clusters by incorporating structural/biological features such as proximity of mutations to functional sites (active sites, phosphorylation sites, disulfide bonds, etc.), solvent accessibility, physiochemical property change of mutations, free energy change of mutations, conservation of residue sites, secondary structure state of residue sites, and expression/phosphorylation changes of samples containing mutations. We have curated experimentally validated mutations identified as neutral or oncogenic from various databases to serve as our training sets. This algorithm can be trained on the curated mutations in various protein subclasses such as homologous proteins, oncogenes, tumor suppressors, etc. to identify distinct structural feature signatures per subclass specific to driver mutations. This tool will aid in revealing putative driver mutations in genes not previously linked with cancer and help pinpoint mutations in known cancer genes that are driving cancer. Specifically, we are interested in applying the algorithm to druggable protein families such as G-Protein Coupled Receptors, Kinases, and Nuclear Hormone Receptors to better understand their role in tumor initiation and progression. Citation Format: Sohini Sengupta, Adam Scott, Amila Weerasinghe, Dan C. Zhou, Matthew A. Wyczalkowski, Reyka G. Jayasinghe, Ken Chen, Gordon Mills, Mike C. Wendl, John Dipersio, Li Ding. Utilizing biological and protein structure-guided features to improve driver mutation discovery [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2357.