PanEffect: a pan-genome visualization tool for variant effects in maize.

SUMMARY: Understanding the effects of genetic variants is crucial for accurately predicting traits and functional outcomes. Recent approaches have utilized artificial intelligence and protein language models to score all possible missense variant effects at the proteome level for a single genome, but a reliable tool is needed to explore these effects at the pan-genome level. To address this gap, we introduce a new tool called PanEffect. We implemented PanEffect at MaizeGDB to enable a comprehensive examination of the potential effects of coding variants across 50 maize genomes. The tool allows users to visualize over 550 million possible amino acid substitutions in the B73 maize reference genome and to observe the effects of the 2.3 million natural variations in the maize pan-genome. Each variant effect score, calculated from the Evolutionary Scale Modeling (ESM) protein language model, shows the log-likelihood ratio difference between B73 and all variants in the pan-genome. These scores are shown using heatmaps spanning benign outcomes to potential functional consequences. In addition, PanEffect displays secondary structures and functional domains along with the variant effects, offering additional functional and structural context. Using PanEffect, researchers now have a platform to explore protein variants and identify genetic targets for crop enhancement. AVAILABILITY AND IMPLEMENTATION: The PanEffect code is freely available on GitHub (https://github.com/Maize-Genetics-and-Genomics-Database/PanEffect). A maize implementation of PanEffect and underlying datasets are available at MaizeGDB (https://www.maizegdb.org/effect/maize/).

Functional annotation and meta-analysis of maize transcriptomes reveal genes involved in biotic and abiotic stress.

BACKGROUND: Environmental stress factors, such as biotic and abiotic stress, are becoming more common due to climate variability, significantly affecting global maize yield. Transcriptome profiling studies provide insights into the molecular mechanisms underlying stress response in maize, though the functions of many genes are still unknown. To enhance the functional annotation of maize-specific genes, MaizeGDB has outlined a data-driven approach with an emphasis on identifying genes and traits related to biotic and abiotic stress. RESULTS: We mapped high-quality RNA-Seq expression reads from 24 different publicly available datasets (17 abiotic and seven biotic studies) generated from the B73 cultivar to the recent version of the reference genome B73 (B73v5) and deduced stress-related functional annotation of maize gene models. We conducted a robust meta-analysis of the transcriptome profiles from the datasets to identify maize loci responsive to stress, identifying 3,230 differentially expressed genes (DEGs): 2,555 DEGs regulated in response to abiotic stress, 408 DEGs regulated during biotic stress, and 267 common DEGs (co-DEGs) that overlap between abiotic and biotic stress. We discovered hub genes from network analyses, and among the hub genes of the co-DEGs we identified a putative NAC domain transcription factor superfamily protein (Zm00001eb369060) IDP275, which previously responded to herbivory and drought stress. IDP275 was up-regulated in our analysis in response to eight different abiotic and four different biotic stresses. A gene set enrichment and pathway analysis of hub genes of the co-DEGs revealed hormone-mediated signaling processes and phenylpropanoid biosynthesis pathways, respectively. Using phylostratigraphic analysis, we also demonstrated how abiotic and biotic stress genes differentially evolve to adapt to changing environments. CONCLUSIONS: These results will help facilitate the functional annotation of multiple stress response gene models and annotation in maize. Data can be accessed and downloaded at the Maize Genetics and Genomics Database (MaizeGDB).

Co-expression pan-network reveals genes involved in complex traits within maize pan-genome.

BACKGROUND: With the advances in the high throughput next generation sequencing technologies, genome-wide association studies (GWAS) have identified a large set of variants associated with complex phenotypic traits at a very fine scale. Despite the progress in GWAS, identification of genotype-phenotype relationship remains challenging in maize due to its nature with dozens of variants controlling the same trait. As the causal variations results in the change in expression, gene expression analyses carry a pivotal role in unraveling the transcriptional regulatory mechanisms behind the phenotypes. RESULTS: To address these challenges, we incorporated the gene expression and GWAS-driven traits to extend the knowledge of genotype-phenotype relationships and transcriptional regulatory mechanisms behind the phenotypes. We constructed a large collection of gene co-expression networks and identified more than 2 million co-expressing gene pairs in the GWAS-driven pan-network which contains all the gene-pairs in individual genomes of the nested association mapping (NAM) population. We defined four sub-categories for the pan-network: (1) core-network contains the highest represented ~ 1% of the gene-pairs, (2) near-core network contains the next highest represented 1-5% of the gene-pairs, (3) private-network contains ~ 50% of the gene pairs that are unique to individual genomes, and (4) the dispensable-network contains the remaining 50-95% of the gene-pairs in the maize pan-genome. Strikingly, the private-network contained almost all the genes in the pan-network but lacked half of the interactions. We performed gene ontology (GO) enrichment analysis for the pan-, core-, and private- networks and compared the contributions of variants overlapping with genes and promoters to the GWAS-driven pan-network. CONCLUSIONS: Gene co-expression networks revealed meaningful information about groups of co-regulated genes that play a central role in regulatory processes. Pan-network approach enabled us to visualize the global view of the gene regulatory network for the studied system that could not be well inferred by the core-network alone.

qTeller: a tool for comparative multi-genomic gene expression analysis.

MOTIVATION: Over the last decade, RNA-Seq whole-genome sequencing has become a widely used method for measuring and understanding transcriptome-level changes in gene expression. Since RNA-Seq is relatively inexpensive, it can be used on multiple genomes to evaluate gene expression across many different conditions, tissues and cell types. Although many tools exist to map and compare RNA-Seq at the genomics level, few web-based tools are dedicated to making data generated for individual genomic analysis accessible and reusable at a gene-level scale for comparative analysis between genes, across different genomes and meta-analyses. RESULTS: To address this challenge, we revamped the comparative gene expression tool qTeller to take advantage of the growing number of public RNA-Seq datasets. qTeller allows users to evaluate gene expression data in a defined genomic interval and also perform two-gene comparisons across multiple user-chosen tissues. Though previously unpublished, qTeller has been cited extensively in the scientific literature, demonstrating its importance to researchers. Our new version of qTeller now supports multiple genomes for intergenomic comparisons, and includes capabilities for both mRNA and protein abundance datasets. Other new features include support for additional data formats, modernized interface and back-end database and an optimized framework for adoption by other organisms' databases. AVAILABILITY AND IMPLEMENTATION: The source code for qTeller is open-source and available through GitHub (https://github.com/Maize-Genetics-and-Genomics-Database/qTeller). A maize instance of qTeller is available at the Maize Genetics and Genomics database (MaizeGDB) (https://qteller.maizegdb.org/), where we have mapped over 200 unique datasets from GenBank across 27 maize genomes. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

FINDER: an automated software package to annotate eukaryotic genes from RNA-Seq data and associated protein sequences.

BACKGROUND: Gene annotation in eukaryotes is a non-trivial task that requires meticulous analysis of accumulated transcript data. Challenges include transcriptionally active regions of the genome that contain overlapping genes, genes that produce numerous transcripts, transposable elements and numerous diverse sequence repeats. Currently available gene annotation software applications depend on pre-constructed full-length gene sequence assemblies which are not guaranteed to be error-free. The origins of these sequences are often uncertain, making it difficult to identify and rectify errors in them. This hinders the creation of an accurate and holistic representation of the transcriptomic landscape across multiple tissue types and experimental conditions. Therefore, to gauge the extent of diversity in gene structures, a comprehensive analysis of genome-wide expression data is imperative. RESULTS: We present FINDER, a fully automated computational tool that optimizes the entire process of annotating genes and transcript structures. Unlike current state-of-the-art pipelines, FINDER automates the RNA-Seq pre-processing step by working directly with raw sequence reads and optimizes gene prediction from BRAKER2 by supplementing these reads with associated proteins. The FINDER pipeline (1) reports transcripts and recognizes genes that are expressed under specific conditions, (2) generates all possible alternatively spliced transcripts from expressed RNA-Seq data, (3) analyzes read coverage patterns to modify existing transcript models and create new ones, and (4) scores genes as high- or low-confidence based on the available evidence across multiple datasets. We demonstrate the ability of FINDER to automatically annotate a diverse pool of genomes from eight species. CONCLUSIONS: FINDER takes a completely automated approach to annotate genes directly from raw expression data. It is capable of processing eukaryotic genomes of all sizes and requires no manual supervision-ideal for bench researchers with limited experience in handling computational tools.

A pan-genomic approach to genome databases using maize as a model system.

Research in the past decade has demonstrated that a single reference genome is not representative of a species' diversity. MaizeGDB introduces a pan-genomic approach to hosting genomic data, leveraging the large number of diverse maize genomes and their associated datasets to quickly and efficiently connect genomes, gene models, expression, epigenome, sequence variation, structural variation, transposable elements, and diversity data across genomes so that researchers can easily track the structural and functional differences of a locus and its orthologs across maize. We believe our framework is unique and provides a template for any genomic database poised to host large-scale pan-genomic data.

MaizeGDB 2018: the maize multi-genome genetics and genomics database.

Since its 2015 update, MaizeGDB, the Maize Genetics and Genomics database, has expanded to support the sequenced genomes of many maize inbred lines in addition to the B73 reference genome assembly. Curation and development efforts have targeted high quality datasets and tools to support maize trait analysis, germplasm analysis, genetic studies, and breeding. MaizeGDB hosts a wide range of data including recent support of new data types including genome metadata, RNA-seq, proteomics, synteny, and large-scale diversity. To improve access and visualization of data types several new tools have been implemented to: access large-scale maize diversity data (SNPversity), download and compare gene expression data (qTeller), visualize pedigree data (Pedigree Viewer), link genes with phenotype images (MaizeDIG), and enable flexible user-specified queries to the MaizeGDB database (MaizeMine). MaizeGDB also continues to be the community hub for maize research, coordinating activities and providing technical support to the maize research community. Here we report the changes MaizeGDB has made within the last three years to keep pace with recent software and research advances, as well as the pan-genomic landscape that cheaper and better sequencing technologies have made possible. MaizeGDB is accessible online at https://www.maizegdb.org.

GenomeQC: a quality assessment tool for genome assemblies and gene structure annotations.

BACKGROUND: Genome assemblies are foundational for understanding the biology of a species. They provide a physical framework for mapping additional sequences, thereby enabling characterization of, for example, genomic diversity and differences in gene expression across individuals and tissue types. Quality metrics for genome assemblies gauge both the completeness and contiguity of an assembly and help provide confidence in downstream biological insights. To compare quality across multiple assemblies, a set of common metrics are typically calculated and then compared to one or more gold standard reference genomes. While several tools exist for calculating individual metrics, applications providing comprehensive evaluations of multiple assembly features are, perhaps surprisingly, lacking. Here, we describe a new toolkit that integrates multiple metrics to characterize both assembly and gene annotation quality in a way that enables comparison across multiple assemblies and assembly types. RESULTS: Our application, named GenomeQC, is an easy-to-use and interactive web framework that integrates various quantitative measures to characterize genome assemblies and annotations. GenomeQC provides researchers with a comprehensive summary of these statistics and allows for benchmarking against gold standard reference assemblies. CONCLUSIONS: The GenomeQC web application is implemented in R/Shiny version 1.5.9 and Python 3.6 and is freely available at https://genomeqc.maizegdb.org/ under the GPL license. All source code and a containerized version of the GenomeQC pipeline is available in the GitHub repository https://github.com/HuffordLab/GenomeQC.

Tissue-specific gene expression and protein abundance patterns are associated with fractionation bias in maize.

BACKGROUND: Maize experienced a whole-genome duplication event approximately 5 to 12 million years ago. Because this event occurred after speciation from sorghum, the pre-duplication subgenomes can be partially reconstructed by mapping syntenic regions to the sorghum chromosomes. During evolution, maize has had uneven gene loss between each ancient subgenome. Fractionation and divergence between these genomes continue today, constantly changing genetic make-up and phenotypes and influencing agronomic traits. RESULTS: Here we regenerate the subgenome reconstructions for the most recent maize reference genome assembly. Based on both expression and abundance data for homeologous gene pairs across multiple tissues, we observed functional divergence of genes across subgenomes. Although the genes in the larger maize subgenome are often expressing more highly than their homeologs in the smaller subgenome, we observed cases where homeolog expression dominance switches in different tissues. We demonstrate for the first time that protein abundances are higher in the larger subgenome, but they also show tissue-specific dominance, a pattern similar to RNA expression dominance. We also find that pollen expression is uniquely decoupled from protein abundance. CONCLUSION: Our study shows that the larger subgenome has a greater range of functional assignments and that there is a relative lack of overlap between the subgenomes in terms of gene functions than would be suggested by similar patterns of gene expression and protein abundance. Our study also revealed that some reactions are catalyzed uniquely by the larger and smaller subgenomes. The tissue-specific, nonequivalent expression-level dominance pattern observed here implies a change in regulatory control which favors differentiated selective pressure on the retained duplicates leading to eventual change in gene functions.

PedigreeNet: a web-based pedigree viewer for biological databases.

MOTIVATION: Plant breeding aims to improve current germplasm that can tolerate a wide range of biotic and abiotic stresses. To accomplish this goal, breeders rely on developing a deeper understanding of genetic makeup and relationships between plant varieties to make informed plant selections. Although rapid advances in genotyping technology generated a large amount of data for breeders, tools that facilitate pedigree analysis and visualization are scant, leaving breeders to use classical, but inherently limited, hierarchical pedigree diagrams for a handful of plant varieties. To answer this need, we developed a simple web-based tool that can be easily implemented at biological databases, called PedigreeNet, to create and visualize customizable pedigree relationships in a network context, displaying pre- and user-uploaded data. RESULTS: As a proof-of-concept, we implemented PedigreeNet at the maize model organism database, MaizeGDB. The PedigreeNet viewer at MaizeGDB has a dynamically-generated pedigree network of 4706 maize lines and 5487 relationships that are currently available as both a stand-alone web-based tool and integrated directly on the MaizeGDB Stock Pages. The tool allows the user to apply a number of filters, select or upload their own breeding relationships, center a pedigree network on a plant variety, identify the common ancestor between two varieties, and display the shortest path(s) between two varieties on the pedigree network. The PedigreeNet code layer is written as a JavaScript wrapper around Cytoscape Web. PedigreeNet fills a great need for breeders to have access to an online tool to represent and visually customize pedigree relationships. AVAILABILITY AND IMPLEMENTATION: PedigreeNet is accessible at https://www.maizegdb.org/breeders_toolbox. The open source code is publically and freely available at GitHub: https://github.com/Maize-Genetics-and-Genomics-Database/PedigreeNet. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Haplotype structure in commercial maize breeding programs in relation to key founder lines.

KEY MESSAGE: High-density haplotype analysis revealed significant haplotype sharing between ex-PVPs registered from 1976 to 1992 and key maize founders, and uncovered similarities and differences in haplotype sharing patterns by company and heterotic group. Proprietary inbreds developed by the private seed industry have been the major source for driving genetic gain in successful North American maize hybrids for decades. Much of the history of industry germplasm can be traced back to key founder lines, some of which were pivotal in the development of prominent heterotic groups. Previous studies have summarized pedigree-based relationships, genetic diversity and population structure among commercial inbreds with expired Plant Variety Protection (ex-PVP). However, less is known about the extent of haplotype sharing between historical founders and ex-PVPs. A better understanding of the relationships between founders and ex-PVPs provides insight into the haplotype and heterotic group structure among industry germplasm. We performed high-density haplotype analysis with 11.3 million SNPs on 212 maize inbreds, which included 157 ex-PVPs registered 1976-1992 and 55 public lines relevant to PVPs. Among these lines were 12 key founders identified in literature review: 207, A632, B14, B37, B73, LH123HT, LH82, Mo17, Oh43, OH7, PHG39 and Wf9. Our results revealed that, on average, 81.6% of an ex-PVP's genome is shared with at least 1 of these 12 founder lines and more than half when limited to B73, Mo17 and 207. Quantifiable similarities and contrasts among heterotic groups and major US seed industry companies were also observed. The results from this study provide high-resolution haplotype data on ex-PVP germplasm, confirm founder relationship trends observed in previous studies, uncover region-specific haplotype structure differences and demonstrate how haplotype sharing analysis can be used as a tool to explore germplasm diversity.

Crowdsourcing image analysis for plant phenomics to generate ground truth data for machine learning.

The accuracy of machine learning tasks critically depends on high quality ground truth data. Therefore, in many cases, producing good ground truth data typically involves trained professionals; however, this can be costly in time, effort, and money. Here we explore the use of crowdsourcing to generate a large number of training data of good quality. We explore an image analysis task involving the segmentation of corn tassels from images taken in a field setting. We investigate the accuracy, speed and other quality metrics when this task is performed by students for academic credit, Amazon MTurk workers, and Master Amazon MTurk workers. We conclude that the Amazon MTurk and Master Mturk workers perform significantly better than the for-credit students, but with no significant difference between the two MTurk worker types. Furthermore, the quality of the segmentation produced by Amazon MTurk workers rivals that of an expert worker. We provide best practices to assess the quality of ground truth data, and to compare data quality produced by different sources. We conclude that properly managed crowdsourcing can be used to establish large volumes of viable ground truth data at a low cost and high quality, especially in the context of high throughput plant phenotyping. We also provide several metrics for assessing the quality of the generated datasets.

MaizeGDB update: new tools, data and interface for the maize model organism database.

MaizeGDB is a highly curated, community-oriented database and informatics service to researchers focused on the crop plant and model organism Zea mays ssp. mays. Although some form of the maize community database has existed over the last 25 years, there have only been two major releases. In 1991, the original maize genetics database MaizeDB was created. In 2003, the combined contents of MaizeDB and the sequence data from ZmDB were made accessible as a single resource named MaizeGDB. Over the next decade, MaizeGDB became more sequence driven while still maintaining traditional maize genetics datasets. This enabled the project to meet the continued growing and evolving needs of the maize research community, yet the interface and underlying infrastructure remained unchanged. In 2015, the MaizeGDB team completed a multi-year effort to update the MaizeGDB resource by reorganizing existing data, upgrading hardware and infrastructure, creating new tools, incorporating new data types (including diversity data, expression data, gene models, and metabolic pathways), and developing and deploying a modern interface. In addition to coordinating a data resource, the MaizeGDB team coordinates activities and provides technical support to the maize research community. MaizeGDB is accessible online at http://www.maizegdb.org.

Automated update, revision, and quality control of the maize genome annotations using MAKER-P improves the B73 RefGen_v3 gene models and identifies new genes.

The large size and relative complexity of many plant genomes make creation, quality control, and dissemination of high-quality gene structure annotations challenging. In response, we have developed MAKER-P, a fast and easy-to-use genome annotation engine for plants. Here, we report the use of MAKER-P to update and revise the maize (Zea mays) B73 RefGen_v3 annotation build (5b+) in less than 3 h using the iPlant Cyberinfrastructure. MAKER-P identified and annotated 4,466 additional, well-supported protein-coding genes not present in the 5b+ annotation build, added additional untranslated regions to 1,393 5b+ gene models, identified 2,647 5b+ gene models that lack any supporting evidence (despite the use of large and diverse evidence data sets), identified 104,215 pseudogene fragments, and created an additional 2,522 noncoding gene annotations. We also describe a method for de novo training of MAKER-P for the annotation of newly sequenced grass genomes. Collectively, these results lead to the 6a maize genome annotation and demonstrate the utility of MAKER-P for rapid annotation, management, and quality control of grasses and other difficult-to-annotate plant genomes.

Harnessing the predicted maize pan-interactome for putative gene function prediction and prioritization of candidate genes for important traits.

The recent assembly and annotation of the 26 maize nested association mapping population founder inbreds have enabled large-scale pan-genomic comparative studies. These studies have expanded our understanding of agronomically important traits by integrating pan-transcriptomic data with trait-specific gene candidates from previous association mapping results. In contrast to the availability of pan-transcriptomic data, obtaining reliable protein-protein interaction (PPI) data has remained a challenge due to its high cost and complexity. We generated predicted PPI networks for each of the 26 genomes using the established STRING database. The individual genome-interactomes were then integrated to generate core- and pan-interactomes. We deployed the PPI clustering algorithm ClusterONE to identify numerous PPI clusters that were functionally annotated using gene ontology (GO) functional enrichment, demonstrating a diverse range of enriched GO terms across different clusters. Additional cluster annotations were generated by integrating gene coexpression data and gene description annotations, providing additional useful information. We show that the functionally annotated PPI clusters establish a useful framework for protein function prediction and prioritization of candidate genes of interest. Our study not only provides a comprehensive resource of predicted PPI networks for 26 maize genomes but also offers annotated interactome clusters for predicting protein functions and prioritizing gene candidates. The source code for the Python implementation of the analysis workflow and a standalone web application for accessing the analysis results are available at https://github.com/eporetsky/PanPPI.

Enhanced pan-genomic resources at the maize genetics and genomics database.

Pan-genomes, encompassing the entirety of genetic sequences found in a collection of genomes within a clade, are more useful than single reference genomes for studying species diversity. This is especially true for a species like Zea mays, which has a particularly diverse and complex genome. Presenting pan-genome data, analyses, and visualization is challenging, especially for a diverse species, but more so when pan-genomic data is linked to extensive gene model and gene data, including classical gene information, markers, insertions, expression and proteomic data, and protein structures as is the case at MaizeGDB. Here, we describe MaizeGDB's expansion to include the genic subset of the Zea pan-genome in a pan-gene data center featuring the maize genomes hosted at MaizeGDB, and the outgroup teosinte Zea genomes from the Pan-Andropoganeae project. The new data center offers a variety of browsing and visualization tools, including sequence alignment visualization, gene trees and other tools, to explore pan-genes in Zea that were calculated by the pipeline Pandagma. Combined, these data will help maize researchers study the complexity and diversity of Zea, and to use the comparative functions to validate pan-gene relationships for a selected gene model.

PhosBoost: Improved phosphorylation prediction recall using gradient boosting and protein language models.

Protein phosphorylation is a dynamic and reversible post-translational modification that regulates a variety of essential biological processes. The regulatory role of phosphorylation in cellular signaling pathways, protein-protein interactions, and enzymatic activities has motivated extensive research efforts to understand its functional implications. Experimental protein phosphorylation data in plants remains limited to a few species, necessitating a scalable and accurate prediction method. Here, we present PhosBoost, a machine-learning approach that leverages protein language models and gradient-boosting trees to predict protein phosphorylation from experimentally derived data. Trained on data obtained from a comprehensive plant phosphorylation database, qPTMplants, we compared the performance of PhosBoost to existing protein phosphorylation prediction methods, PhosphoLingo and DeepPhos. For serine and threonine prediction, PhosBoost achieved higher recall than PhosphoLingo and DeepPhos (.78, .56, and .14, respectively) while maintaining a competitive area under the precision-recall curve (.54, .56, and .42, respectively). PhosphoLingo and DeepPhos failed to predict any tyrosine phosphorylation sites, while PhosBoost achieved a recall score of .6. Despite the precision-recall tradeoff, PhosBoost offers improved performance when recall is prioritized while consistently providing more confident probability scores. A sequence-based pairwise alignment step improved prediction results for all classifiers by effectively increasing the number of inferred positive phosphosites. We provide evidence to show that PhosBoost models are transferable across species and scalable for genome-wide protein phosphorylation predictions. PhosBoost is freely and publicly available on GitHub.

Maize Feature Store: A centralized resource to manage and analyze curated maize multi-omics features for machine learning applications.

The big-data analysis of complex data associated with maize genomes accelerates genetic research and improves agronomic traits. As a result, efforts have increased to integrate diverse datasets and extract meaning from these measurements. Machine learning models are a powerful tool for gaining knowledge from large and complex datasets. However, these models must be trained on high-quality features to succeed. Currently, there are no solutions to host maize multi-omics datasets with end-to-end solutions for evaluating and linking features to target gene annotations. Our work presents the Maize Feature Store (MFS), a versatile application that combines features built on complex data to facilitate exploration, modeling and analysis. Feature stores allow researchers to rapidly deploy machine learning applications by managing and providing access to frequently used features. We populated the MFS for the maize reference genome with over 14 000 gene-based features based on published genomic, transcriptomic, epigenomic, variomic and proteomics datasets. Using the MFS, we created an accurate pan-genome classification model with an AUC-ROC score of 0.87. The MFS is publicly available through the maize genetics and genomics database. Database URL  https://mfs.maizegdb.org/.

Maize protein structure resources at the maize genetics and genomics database.

Protein structures play an important role in bioinformatics, such as in predicting gene function or validating gene model annotation. However, determining protein structure was, until now, costly and time-consuming, which resulted in a structural biology bottleneck. With the release of such programs AlphaFold and ESMFold, this bottleneck has been reduced by several orders of magnitude, permitting protein structural comparisons of entire genomes within reasonable timeframes. MaizeGDB has leveraged this technological breakthrough by offering several new tools to accelerate protein structural comparisons between maize and other plants as well as human and yeast outgroups. MaizeGDB also offers bulk downloads of these comparative protein structure data, along with predicted functional annotation information. In this way, MaizeGDB is poised to assist maize researchers in assessing functional homology, gene model annotation quality, and other information unavailable to maize scientists even a few years ago.

Predicting Tissue-Specific mRNA and Protein Abundance in Maize: A Machine Learning Approach.

Machine learning and modeling approaches have been used to classify protein sequences for a broad set of tasks including predicting protein function, structure, expression, and localization. Some recent studies have successfully predicted whether a given gene is expressed as mRNA or even translated to proteins potentially, but given that not all genes are expressed in every condition and tissue, the challenge remains to predict condition-specific expression. To address this gap, we developed a machine learning approach to predict tissue-specific gene expression across 23 different tissues in maize, solely based on DNA promoter and protein sequences. For class labels, we defined high and low expression levels for mRNA and protein abundance and optimized classifiers by systematically exploring various methods and combinations of k-mer sequences in a two-phase approach. In the first phase, we developed Markov model classifiers for each tissue and built a feature vector based on the predictions. In the second phase, the feature vector was used as an input to a Bayesian network for final classification. Our results show that these methods can achieve high classification accuracy of up to 95% for predicting gene expression for individual tissues. By relying on sequence alone, our method works in settings where costly experimental data are unavailable and reveals useful insights into the functional, evolutionary, and regulatory characteristics of genes.