Taxonomic classification of DNA sequences beyond sequence similarity using deep neural networks.

Taxonomic classification, that is, the assignment to biological clades with shared ancestry, is a common task in genetics, mainly based on a genome similarity search of large genome databases. The classification quality depends heavily on the database, since representative relatives must be present. Many genomic sequences cannot be classified at all or only with a high misclassification rate. Here we present BERTax, a deep neural network program based on natural language processing to precisely classify the superkingdom and phylum of DNA sequences taxonomically without the need for a known representative relative from a database. We show BERTax to be at least on par with the state-of-the-art approaches when taxonomically similar species are part of the training data. For novel organisms, however, BERTax clearly outperforms any existing approach. Finally, we show that BERTax can also be combined with database approaches to further increase the prediction quality in almost all cases. Since BERTax is not based on similar entries in databases, it allows precise taxonomic classification of a broader range of genomic sequences, thus increasing the overall information gain.

NGS read classification using AI.

Clinical metagenomics is a powerful diagnostic tool, as it offers an open view into all DNA in a patient's sample. This allows the detection of pathogens that would slip through the cracks of classical specific assays. However, due to this unspecific nature of metagenomic sequencing, a huge amount of unspecific data is generated during the sequencing itself and the diagnosis only takes place at the data analysis stage where relevant sequences are filtered out. Typically, this is done by comparison to reference databases. While this approach has been optimized over the past years and works well to detect pathogens that are represented in the used databases, a common challenge in analysing a metagenomic patient sample arises when no pathogen sequences are found: How to determine whether truly no evidence of a pathogen is present in the data or whether the pathogen's genome is simply absent from the database and the sequences in the dataset could thus not be classified? Here, we present a novel approach to this problem of detecting novel pathogens in metagenomic datasets by classifying the (segments of) proteins encoded by the sequences in the datasets. We train a neural network on the sequences of coding sequences, labeled by taxonomic domain, and use this neural network to predict the taxonomic classification of sequences that can not be classified by comparison to a reference database, thus facilitating the detection of potential novel pathogens.

Comparative analysis of chicken cecal microbial diversity and taxonomic composition in response to dietary variation using 16S rRNA amplicon sequencing.

BACKGROUND: Antibiotic resistance poses a grave threat to One-Health. By replacing antibiotics with non-antibiotic additives (are alternatives to antibiotics, ATAs) like phytogenic feed additives and organic acids in poultry feed. ATAs are a potential alternative as these decline the proliferation of pathogenic bacteria and strengthen gut function in broiler chickens. In this study, we use 16S rRNA amplicon sequencing of the V3-V4 region to evaluate phytogenic feed additives and organic acids on the cecal microbial diversity of broiler chickens. METHODS AND RESULTS: Two hundred & forty broiler chicks were divided into five treatments comprising: a controlled basal diet (CON), antibiotic group (AB), phytogenic feed additives (PHY), organic acids (ORG), and a combination of PHY + ORG (COM). A distinctive microbial community structure was observed amongst different treatments with increased microbial diversity in AB, ORG, and COM (p < 0.05). The synergistic effects of PHY and ORG increased bacterial population of phyla: Firmicutes, Bacteroides, and Proteobacteria in the cecum. The presence of species, Akkermansia muciniphila (involved in mucin degradation) and Bacillus safensis (a probiotic bacterium) were noticed in COM and PHY, respectively. Clustering analysis revealed a higher relative abundance of similar microbial community composition between AB and ORG groups. CONCLUSIONS: Treatments with PHY and ORG modified the relative abundance and presence/absence of specific microbiota in the chicken cecum. Hence, cecal microbiota modulation through diet is a promising strategy to reduce cross-contamination of zoonotic poultry pathogens, led to healthy and economical broiler meat.

Robust direct digital-to-biological data storage in living cells.

DNA has been the predominant information storage medium for biology and holds great promise as a next-generation high-density data medium in the digital era. Currently, the vast majority of DNA-based data storage approaches rely on in vitro DNA synthesis. As such, there are limited methods to encode digital data into the chromosomes of living cells in a single step. Here, we describe a new electrogenetic framework for direct storage of digital data in living cells. Using an engineered redox-responsive CRISPR adaptation system, we encoded binary data in 3-bit units into CRISPR arrays of bacterial cells by electrical stimulation. We demonstrate multiplex data encoding into barcoded cell populations to yield meaningful information storage and capacity up to 72 bits, which can be maintained over many generations in natural open environments. This work establishes a direct digital-to-biological data storage framework and advances our capacity for information exchange between silicon- and carbon-based entities.

Predictors of maternal-origin microchimerism in young women in the Philippines.

OBJECTIVES: Microchimerism is the presence of a small quantity of cells or DNA from a genetically distinct individual. This phenomenon occurs with bidirectional maternal-fetal exchange during pregnancy. Microchimerism can persist for decades after delivery and have long-term health implications. However, little is known about why microchimerism is detectable at varying levels in different individuals. We examine the variability and the following potential determinants of maternal-origin microchimerism (MMc) in young women in the Philippines: gestational duration (in utero exposure to MMc), history of being breastfed (postpartum exposure to MMc), maternal telomere length (maternal cells' ability to replicate and persist), and participant's pregnancies in young adulthood (effect of adding fetal-origin microchimerism to preexisting MMc). MATERIALS AND METHODS: Data are from the Cebu Longitudinal Health and Nutrition Survey, a population-based study of infant feeding practices and long-term health outcomes. We quantified MMc using quantitative PCR (qPCR) in 89 female participants, ages 20-22, and analyzed these data using negative binomial regression. RESULTS: In a multivariate model including all predictors, being breastfed substantially predicted decreased MMc (detection rate ratio = 0.15, p = 0.007), and there was a trend of decreasing MMc in participants who had experienced more pregnancies (detection rate ratio = 0.55, p = 0.057). DISCUSSION: These results might be explained by breastfeeding having lasting impact on immune regulatory networks, thus reducing MMc persistence. MMc may also decrease in response to the introduction of fetal-origin microchimerism with pregnancies experienced in adulthood.

Definition and Usage of Texture Feature for Biological Sequence.

In recent years, sequencing technology has developed rapidly. This produces a large number of biological sequence data. Because of its importance, there have been many studies on biological sequences. However, there is still a lack of an effective quantitative method for defining and calculating texture features of biological sequences. Texture is an important visual feature. It is generally used to describe the spatial arrangement of intensities of images. Here we defined the texture features of biological sequence. Combining the digital coding of biological sequence with the calculation method of image texture features, we defined the texture features of biological sequence and designed the calculation method. We applied this method to DNA sequence features quantification and analysis. Using these quantified features, we can compute the similarity distance matrix of DNA sequences and construct the phylogenetic relationships based on the clustering of the quantified features. This method can be applied to analyze any biological sequence, and all biological sequences can be digitally coded and texture features can be calculated by this method. This is a novel study of biological sequence texture features. This will usher in a new era of quantitative and mathematical calculation of biological sequence features.

Identification of protein-protected mRNA fragments and structured excised intron RNAs in human plasma by TGIRT-seq peak calling.

Human plasma contains > 40,000 different coding and non-coding RNAs that are potential biomarkers for human diseases. Here, we used thermostable group II intron reverse transcriptase sequencing (TGIRT-seq) combined with peak calling to simultaneously profile all RNA biotypes in apheresis-prepared human plasma pooled from healthy individuals. Extending previous TGIRT-seq analysis, we found that human plasma contains largely fragmented mRNAs from > 19,000 protein-coding genes, abundant full-length, mature tRNAs and other structured small non-coding RNAs, and less abundant tRNA fragments and mature and pre-miRNAs. Many of the mRNA fragments identified by peak calling correspond to annotated protein-binding sites and/or have stable predicted secondary structures that could afford protection from plasma nucleases. Peak calling also identified novel repeat RNAs, miRNA-sized RNAs, and putatively structured intron RNAs of potential biological, evolutionary, and biomarker significance, including a family of full-length excised intron RNAs, subsets of which correspond to mirtron pre-miRNAs or agotrons.

Evaluation of methods for detecting human reads in microbial sequencing datasets.

Sequencing data from host-associated microbes can often be contaminated by the body of the investigator or research subject. Human DNA is typically removed from microbial reads either by subtractive alignment (dropping all reads that map to the human genome) or by using a read classification tool to predict those of human origin, and then discarding them. To inform best practice guidelines, we benchmarked eight alignment-based and two classification-based methods of human read detection using simulated data from 10 clinically prevalent bacteria and three viruses, into which contaminating human reads had been added. While the majority of methods successfully detected >99 % of the human reads, they were distinguishable by variance. The most precise methods, with negligible variance, were Bowtie2 and SNAP, both of which misidentified few, if any, bacterial reads (and no viral reads) as human. While correctly detecting a similar number of human reads, methods based on taxonomic classification, such as Kraken2 and Centrifuge, could misclassify bacterial reads as human, although the extent of this was species-specific. Among the most sensitive methods of human read detection was BWA, although this also made the greatest number of false positive classifications. Across all methods, the set of human reads not identified as such, although often representing <0.1 % of the total reads, were non-randomly distributed along the human genome with many originating from the repeat-rich sex chromosomes. For viral reads and longer (>300 bp) bacterial reads, the highest performing approaches were classification-based, using Kraken2 or Centrifuge. For shorter (c. 150 bp) bacterial reads, combining multiple methods of human read detection maximized the recovery of human reads from contaminated short read datasets without being compromised by false positives. A particularly high-performance approach with shorter bacterial reads was a two-stage classification using Bowtie2 followed by SNAP. Using this approach, we re-examined 11 577 publicly archived bacterial read sets for hitherto undetected human contamination. We were able to extract a sufficient number of reads to call known human SNPs, including those with clinical significance, in 6 % of the samples. These results show that phenotypically distinct human sequence is detectable in publicly archived microbial read datasets.

A unified dinucleotide alphabet describing both RNA and DNA structures.

By analyzing almost 120 000 dinucleotides in over 2000 nonredundant nucleic acid crystal structures, we define 96+1 diNucleotide Conformers, NtCs, which describe the geometry of RNA and DNA dinucleotides. NtC classes are grouped into 15 codes of the structural alphabet CANA (Conformational Alphabet of Nucleic Acids) to simplify symbolic annotation of the prominent structural features of NAs and their intuitive graphical display. The search for nontrivial patterns of NtCs resulted in the identification of several types of RNA loops, some of them observed for the first time. Over 30% of the nearly six million dinucleotides in the PDB cannot be assigned to any NtC class but we demonstrate that up to a half of them can be re-refined with the help of proper refinement targets. A statistical analysis of the preferences of NtCs and CANA codes for the 16 dinucleotide sequences showed that neither the NtC class AA00, which forms the scaffold of RNA structures, nor BB00, the DNA most populated class, are sequence neutral but their distributions are significantly biased. The reported automated assignment of the NtC classes and CANA codes available at dnatco.org provides a powerful tool for unbiased analysis of nucleic acid structures by structural and molecular biologists.

PlasClass improves plasmid sequence classification.

Many bacteria contain plasmids, but separating between contigs that originate on the plasmid and those that are part of the bacterial genome can be difficult. This is especially true in metagenomic assembly, which yields many contigs of unknown origin. Existing tools for classifying sequences of plasmid origin give less reliable results for shorter sequences, are trained using a fraction of the known plasmids, and can be difficult to use in practice. We present PlasClass, a new plasmid classifier. It uses a set of standard classifiers trained on the most current set of known plasmid sequences for different sequence lengths. We tested PlasClass sequence classification on held-out data and simulations, as well as publicly available bacterial isolates and plasmidome samples and plasmids assembled from metagenomic samples. PlasClass outperforms the state-of-the-art plasmid classification tool on shorter sequences, which constitute the majority of assembly contigs, allowing it to achieve higher F1 scores in classifying sequences from a wide range of datasets. PlasClass also uses significantly less time and memory. PlasClass can be used to easily classify plasmid and bacterial genome sequences in metagenomic or isolate assemblies. It is available under the MIT license from: https://github.com/Shamir-Lab/PlasClass.

Fertility care in the era of commercial direct-to-consumer home DNA kits: issues to ponder.

Recently, we have witnessed a shift in the nature of genetic testing. What was once solely in the hands of the scientific community is now easily accessible to anyone. Commercial companies such as Veritas Genetics, Ancestry, and 23andMe offer cheap direct-to-consumer home DNA kits that are branded as a "health and ancestry service." These tests are gaining in popularity, and it is estimated that since their marketing, over 30 million people worldwide have provided their DNA samples. The implications of this new genetic era are diverse, ranging from the individuals personal health assessment, to family genealogy and ancestry, and the complexity of establishing an enormous population-based genetic database. Unique implications of commercial DNA tests on fertility patients and heath care providers, such as in the case of gamete donation, are important to acknowledge and have implications for reproductive care.

Developmental validation of the Microreader™ 20A ID system.

The Microreader™ 20A ID system is designed for forensic applications such as personal identification, parentage testing, and research. It includes 13 combined DNA index system (CODIS) short tandem repeat (STR) loci (CSF1PO, FGA, TH01, TPOX, vWA, D3S1358, D5S818, D7S820, D8S1179, D13S317, D16S539, D18S51, and D21S11), three expanded CODIS STR loci (D12S391, D19S433, and D2S1338), three non-CODIS STR loci (D6S1043, Penta D, and Penta E), and the amelogenin locus in one reaction with a six-dye fluorescent (FAM, HEX, TAMAR, ROX, PUR, and QD550) analysis system. In this study, the Microreader™ 20A ID system was validated according to the Scientific Working Group on DNA Analysis Methods validation guidelines for forensic DNA Analysis methods and Chinese national standard, including PCR-based studies, sensitivity study, precision, and accuracy evaluation, stutter calculation, inhibitor tests, species specificity, and DNA mixture studies. Our results suggest that the Microreader™ 20A ID system is a useful tool for personal identification and parentage testing.

SpCLUST: Towards a fast and reliable clustering for potentially divergent biological sequences.

This paper presents SpCLUST, a new C++ package that takes a list of sequences as input, aligns them with MUSCLE, computes their similarity matrix in parallel and then performs the clustering. SpCLUST extends a previously released software by integrating additional scoring matrices which enables it to cover the clustering of amino-acid sequences. The similarity matrix is now computed in parallel according to the master/slave distributed architecture, using MPI. Performance analysis, realized on two real datasets of 100 nucleotide sequences and 1049 amino-acids ones, show that the resulting library substantially outperforms the original Python package. The proposed package was also intensively evaluated on simulated and real genomic and protein data sets. The clustering results were compared to the most known traditional tools, such as UCLUST, CD-HIT and DNACLUST. The comparison showed that SpCLUST outperforms the other tools when clustering divergent sequences, and contrary to the others, it does not require any user intervention or prior knowledge about the input sequences.

Barcoded DNA nanostructures for the multiplexed profiling of subcellular protein distribution.

Massively parallel DNA sequencing is established, yet high-throughput protein profiling remains challenging. Here, we report a barcoding approach that leverages the combinatorial sequence content and the configurational programmability of DNA nanostructures for high-throughput multiplexed profiling of the subcellular expression and distribution of proteins in whole cells. The barcodes are formed by in situ hybridization of tetrahedral DNA nanostructures and short DNA sequences conjugated with protein-targeting antibodies, and by nanostructure-assisted ligation (either enzymatic or chemical) of the nanostructures and exogenous DNA sequences bound to nanoparticles of different sizes (which cause these localization sequences to differentially distribute across subcellular compartments). Compared with linear DNA barcoding, the nanostructured barcodes enhance the signal by more than 100-fold. By implementing the barcoding approach on a microfluidic device for the analysis of rare patient samples, we show that molecular subtypes of breast cancer can be accurately classified and that subcellular spatial markers of disease aggressiveness can be identified.

Development of a new 25plex STRs typing system for forensic application.

We have developed a novel STR 25-plex florescence multiplex-STR kit (DNATyper25) to genotype 23 autosomal and two sex-linked loci for forensic applications and paternity analysis. Of the 23 autosomal loci, 20 are non-CODIS. The sex-linked markers include a Y-STR locus (DYS391) and the Amelogenin gene. We present developmental validation studies to show that the DNATyper25 kit is reproducible, accurate, sensitive, and robust. Sensitivity testing showed that full profiles were achieved with as low as 125 pg of human DNA. Specificity testing demonstrated a lack of cross reactivity with a variety of commonly encountered non-human DNA contaminants. Stability testing showed that full profiles were obtained with humic acid concentration ≤60 ng/µL and hematin concentration <400 µM. For forensic evaluation, the 23 autosomal STRs followed the Hardy-Weinberg equilibrium. In an analysis of 509 Chinese (CN) Hans, we detected a combined total of 181 alleles at the 23 autosomal STR loci. Since these autosomal STRs are independent from one another, PM was 8.4528 × 10-22 , TDP was 0.999 999 999 999 999 999 999, CEP was 0.999 999 8395. The forensic efficiency parameters demonstrated that these autosomal STRs are highly polymorphic and informative in the Han population of China. We performed population comparisons and showed that the Northern CN Han has a close genetic relationship with the Luzhou Han, Tujia, and Bai populations. We propose that the DNATyper25 kit will be useful for cases where paternity analysis is difficult and for situations where DNA samples are limited in quantity and low in quality.

Development of a novel multiplex polymerase chain reaction system for forensic individual identification using insertion/deletion polymorphisms.

Insertion/deletion (InDel) polymorphisms have been widely used in the fields of population genetics, genetic map constructions, and forensic investigations owing to the advantages of their low mutation rates, widespread distributions in the human genome, and small amplicon sizes. In order to provide more InDels with high discrimination power in Chinese populations, we selected and constructed one novel multiplex PCR-InDel panel for forensic individual identification. Genetic distributions of these 35 InDels in five reference populations from East Asia showed low genetic differentiations among these populations. Forensic efficiency evaluations of these InDels revealed that these loci could perform well for forensic individual identifications in these reference populations. In the meantime, genetic diversities and forensic parameters of these InDels were further investigated in the studied Kazak group. Mean value of polymorphism information content for 35 InDels was 0.3611. Cumulative power of discrimination of 35 InDels was 0.99999999999999603 in Kazak group. Given these results, the panel is suitable for individual identifications in the studied Kazak and these reference populations.

Physicochemical property based computational scheme for classifying DNA sequence elements of Saccharomyces cerevisiae.

GenerationE of huge "omics" data necessitates the development and application of computational methods to annotate the data in terms of biological features. In the context of DNA sequence, it is important to unravel the hidden physicochemical signatures. For this purpose, we have considered various sequence elements such as promoter, ACS, LTRs, telomere, and retrotransposon of the model organism Saccharomyces cerevisiae. Contributions due to di-nucleotides play a major role in studying the DNA conformation profile. The physicochemical parameters used are hydrogen bonding energy, stacking energy and solvation energy per base pair. Our computational study shows that all sequence elements in this study have distinctive physicochemical signatures and the same can be exploited for prediction experiments. The order that we see in a DNA sequence is dictated by biological regions and hence, there exists role of dependency in the sequence makeup, keeping this in mind we are proposing two computational schemes (a) using a windowing block size procedure and (b) using di-nucleotide transitions. We obtained better discriminating profile when we analyzed the sequence data in windowing manner. In the second novel approach, we introduced the di-nucleotide transition probability matrix (DTPM) to study the hidden layer of information embedded in the sequences. DTPM has been used as weights for scanning and predictions. This proposed computational scheme incorporates the memory property which is more realistic to study the physicochemical properties embedded in DNA sequences. Our analysis shows that the DTPM scheme performs better than the existing method in this applied region. Characterization of these elements will be a key to genome editing applications and advanced machine learning approaches may also require such distinctive profiles as useful input features.

Targeting individual cells by barcode in pooled sequence libraries.

Transcriptional profiling of thousands of single cells in parallel by RNA-seq is now routine. However, due to reliance on pooled library preparation, targeting analysis to particular cells of interest is difficult. Here, we present a multiplexed PCR method for targeted sequencing of select cells from pooled single-cell sequence libraries. We demonstrated this molecular enrichment method on multiple cell types within pooled single-cell RNA-seq libraries produced from primary human blood cells. We show how molecular enrichment can be combined with FACS to efficiently target ultra-rare cell types, such as the recently identified AXL+SIGLEC6+ dendritic cell (AS DC) subset, in order to reduce the required sequencing effort to profile single cells by 100-fold. Our results demonstrate that DNA barcodes identifying cells within pooled sequencing libraries can be used as targets to enrich for specific molecules of interest, for example reads from a set of target cells.

Enhancing immune effects of a DNA vaccine against kidney cancer using CD40L as an adjuvant

The use of specific combinations of antigens and adjuvant represents a promising approach for increasing the immunogenicity of DNA vaccines. In the present study, we evaluated the immunity and antitumor effects of DNA vaccines with G250 as the target antigen in a mouse model of renal cell carcinoma. We constructed two recombinant plasmids, pVAX1-G250 and pVAX1-CD40L. The recombinant plasmids were injected into mice by intramuscular injection and electrical pulse stimulation. ELISA and ELISPOT experiments were performed to evaluate the corresponding humoral and cellular immune responses following immunization. To further investigate the antitumor potential of the DNA vaccines, we established a tumor-bearing mouse model expressing G250 target antigen. Our results showed that immunization with the combination of the two plasmids exerted the strongest anti-tumor effects. Therefore, our findings demonstrated the effectiveness of CD40L as an adjuvant for DNA vaccines and highlighted the promising use of these vaccines for the treatment of tumors.

Evolution of potential biomarkers of acute muscle injury after physical exercise

Skeletal muscle injury is a frequent event and diagnosis using the classical blood markers sometimes produces unsatisfactory results. Therefore, objective of the study was to detect new biomarkers in plasma, saliva and urine in response to acute muscle damage induced by physical exercise. A cross-sectional study was conducted with 27 American football players. Before the physical exercises (T0), 60 minutes (T1) and 24 hours (T2) after physical exercise, was determined the clinical, biochemical and molecular parameters, including ADA, TBARS, leukocytes, lymphocytes and comet assay. The serum ADA was significantly higher in T1 and T2, in the urine there was a significant increase in T1, in the saliva there was no significant differences. There was an increase in serum TBARS in T2, saliva and urine in T1. The leukocytes increased in T1 and decreased in T2. Through the comet assay was observed significant DNA damage in T1 and T2. Serum and urinary ADA activity, serum, urinary and salivary TBARS are robust and promising biomarkers of acute muscle injury and that the comet assay allows a quick and effective evaluation of DNA lesions induced by physical exercise and could be used to monitor athletes avoiding injuries that are more serious.