Biology subjects

Zhang, F.

Publications and source records attributed to Zhang, F..

At least 379 records · Page 21Linked to original sources

bioRxiv2020

FASTQuick: Rapid and comprehensive quality assessment of raw sequence reads

BackgroundRapid and thorough quality assessment of sequenced genomes in an ultra-high-throughput scale is crucial for successful large-scale genomic studies. Comprehensive quality assessment typically requires full genome alignment, which costs a significant amount of computational resources and turnaround time. Existing tools are either computational expensive due to full alignment or lacking essential quality metrics by skipping read alignment. FindingsWe developed a set of rapid and accurate methods to produce comprehensive quality metrics directly from raw sequence reads without full genome alignment. Our methods offer orders of magnitude faster turnaround time than existing full alignment-based methods while providing comprehensive and sophisticated quality metrics, including estimates of genetic ancestry and contamination. ConclusionsBy rapidly and comprehensively performing the quality assessment, our tool will help investigators detect potential issues in ultra-high-throughput sequence reads in real-time within a low computational cost, ensuring high-quality downstream analysis and preventing unexpected loss in time, money, and invaluable specimens.

bioinformatics↗

bioRxiv2020

Effective Inhibition of SARS-CoV-2 Entry by Heparin and Enoxaparin Derivatives

Severe acute respiratory syndrome-related coronavirus 2 (SARS-CoV-2) has caused a pandemic of historic proportions and continues to spread globally, with enormous consequences to human health. Currently there is no vaccine, effective therapeutic or prophylactic. Like other betacoronaviruses, attachment and entry of SARS-CoV-2 is mediated by the spike glycoprotein (SGP). In addition to its well-documented interaction with its receptor, human angiotensin converting enzyme 2 (hACE2), SGP has been found to bind to glycosaminoglycans like heparan sulfate, which is found on the surface of virtually all mammalian cells. Here, we pseudotyped SARS-CoV-2 SGP on a third generation lentiviral (pLV) vector and tested the impact of various sulfated polysaccharides on transduction efficiency in mammalian cells. The pLV vector pseudotyped SGP efficiently and produced high titers on HEK293T cells. Various sulfated polysaccharides potently neutralized pLV-S pseudotyped virus with clear structure-based differences in anti-viral activity and affinity to SGP. Concentration-response curves showed that pLV-S particles were efficiently neutralized by a range of concentrations of unfractionated heparin (UFH), enoxaparin, 6-O-desulfated UFH and 6-O-desulfated enoxaparin with an IC50 of 5.99 {micro}g/L, 1.08 mg/L, 1.77 {micro}g/L, and 5.86 mg/L respectively. The low serum bioavailability of intranasally administered UFH, along with data suggesting that the nasal epithelium is a portal for initial infection and transmission, suggest that intranasal administration of UFH may be an effective and safe prophylactic treatment.

microbiology↗

bioRxiv2020

Radiosynthesis and Preclinical Evaluation of Ga-NOTA-Folate for PET Imaging of Folate Receptor β Positive Macrophages

Folate receptor {beta} (FR-{beta}) is one of the markers expressed on macrophages and a promising target for imaging of inflammation. Here, we report the radiosynthesis and preclinical evaluation of [68Ga]Ga-NOTA-folate (68Ga-FOL). First, we determined the affinity of 68Ga-FOL using human FR-{beta} expressing cells. Then, we studied atherosclerotic mice with 68Ga-FOL and 18F-FDG PET/CT. After sacrifice, the tissues excised were measured with a {gamma}-counter for ex vivo biodistribution. Further, the tracer distribution and co-localization with macrophages in aorta cryosections were studied using autoradiography, hematoxylin-eosin staining and immunostaining with anti-Mac-3 antibody. Specificity of 68Ga-FOL was assessed in a blocking study with excess of folate glucosamine. As a last step, human radiation doses were extrapolated from rat PET data. We were able to produce 68Ga-FOL at high radioactivity concentration, with high molar activity and radiochemical purity. The cell binding studies showed high (5.1 {+/-} 1.1 nM) affinity of 68Ga-FOL to FR-{beta}. The myocardial uptake of 68Ga-FOL (SUV 0.43 {+/-} 0.06) was 20-folds lower compared to 18F-FDG (SUV 10.6 {+/-} 1.8, P = 0.001). The autoradiography and immunohistochemistry of aorta revealed that 68Ga-FOL radioactivity co-localized with Mac-3-positive macrophage-rich atherosclerotic plaques. The plaque-to-healthy vessel wall ratio of 68Ga-FOL (2.44 {+/-} 0.15) was significantly higher than that of 18F-FDG (1.93 {+/-} 0.22, P = 0.005). Blocking studies verified 68Ga-FOL specificity to FR. As estimated from rat data the human effective dose was 0.0105 mSv/MBq. The organ with highest absorbed dose was kidney (0.1420 mSv/MBq). In conclusion, 68Ga-FOL is a promising new FR-{beta}-targeted tracer for imaging macrophage-associated inflammation. TABLE OF CONTENT/ABSTRACT GRAPHIC O_FIG_DISPLAY_L [Figure 1] M_FIG_DISPLAY C_FIG_DISPLAY

biochemistry↗

bioRxiv2020

Structure of the unique tetrameric STENOFOLIA homeodomain bound with DNA

Homeobox transcription factors are key regulators of morphogenesis and development in both animals and plants1. In plants, the WUSCHEL-related homeobox (WOX) family transcription factors function as central organizers of several developmental programs from embryo patterning to meristematic stem cell maintenance through transcriptional activation and repression2-4. The structure of WOX Homeodomain (HD) and the molecular mechanism of its interaction with DNA are unknown. Here, we report the 2.1 [A] crystal structure of the STENOFOLIA (STF) HD from Medicago truncatula in complex with DNA. STF binds DNA as a novel cooperative tetramer, enclosing nearly entire bound DNA surface. The STF tetramer is partially stabilized by docking of the C-terminal tail from one protomer onto a conserved hydrophobic surface on the head of another in a head-to-tail manner. Helix 3 not only serves a canonical role as a base reader in the major groove, but also provides extensive binding to DNA in the minor groove. Our structural and functional data reveal that STF specifically targets TGA sequence and the cooperative tetrameric binding with DNA is key to transcriptional repression in plants. Our data reveal an unprecedented HD:DNA recognition mechanism, representing the first plant HD structure from WOX family of transcription factors.

biochemistry↗

bioRxiv2020

The consensus Nglyco-X-S/T motif and a previously unknown Nglyco-N -linked glycosylation are necessary for growth and pathogenicity of Phytophthora

Asparagine (Asn, N) -linked glycosylation within the glycosylation motif (Nglyco-X-S/T; X=P) is a ubiquitously distributed post-translational modification that participates in diverse eukaryotic cellular processes. However, little is known about the characteristic features and roles of N-glycosylation in oomycetes. In this work, it found that 2.5 g/ml tunicamycin (N-glycosylation inhibitor) completely inhibited Phytophthora sojae growth, suggesting that N-glycosylation is necessary for oomycete development. We conducted a glycoproteomic analysis of P. sojae to identify and map all N-glycosylated proteins and to quantify differentially expressed glycoproteins associated with mycelia, asexual cysts, and sexual oospores. A total of 355 N-glycosylated proteins were found, containing 496 glycosites that likely participate in glycan degradation, carbon metabolism, glycolysis, or other central metabolic pathways. To verify the glycoproteomic results and further examine the function of N-glycosylation in P. sojae, two proteins were selected for PNGase F deglycosylation assays and CRISPR/Cas9-mediated site-directed mutagenesis, including a GPI transamidase protein (GPI16) up-regulated in cysts, with the consensus Nglyco-X-S/T motif at Asn 94, and a heat shock protein 70 (HSP70) up-regulated in cysts and oospores with a previously unknown Nglyco-N motif at Asn 270. We demonstrated that the GPI16 and HSP70 are both N-glycosylated proteins, confirming that the Nglyco-N motif is a target site for asparagine - oligosaccharide N-glycosidic linkage. Glycosite mutations of Asn 94 in the GPI16 led to impaired cyst germination and pathogenicity, while HSP70 mutants exhibited decreased cyst germination and oospore production. This work describes an integrated map of oomycete N-glycoproteomes and advances our understanding of N-glycosylation in oomycetes. Moreover, we confirm that the consensus Nglyco-X-S/T and the Nglyco-N -linked glycosites are both essential for the growth of Phytophthora sojae, indicating that there are multiple N-glycosylation motifs in oomycetes.

microbiology↗

bioRxiv2020

CeMbio - The C. elegans microbiome resource

The study of microbiomes by sequencing has revealed a plethora of correlations between microbial community composition and various life-history characteristics of the corresponding host species. However, inferring causation from correlation is often hampered by the sheer compositional complexity of microbiomes, even in simple organisms. Synthetic communities offer an effective approach to infer cause-effect relationships in host-microbiome systems. Yet the available communities suffer from several drawbacks, such as artificial (thus non-natural) choice of microbes, microbe-host mismatch (e.g. human microbes in gnotobiotic mice), or hosts lacking genetic tractability. Here we introduce CeMbio, a simplified natural Caenorhabditis elegans microbiota derived from our previous meta-analysis of the natural microbiome of this nematode. The CeMbio resource is amenable to all strengths of the C. elegans model system, strains included are readily culturable, they all colonize the worm gut individually, and comprise a robust community that distinctly affects nematode life-history. Several tools have additionally been developed for the CeMbio strains, including diagnostic PCR primers, completely sequenced genomes, and metabolic network models. With CeMbio, we provide a versatile resource and toolbox for the in-depth dissection of naturally relevant host-microbiome interactions in C. elegans. Dataset accession numbersWhole genome sequencing data (PRJNA624308); microbiome sequencing [PRJEB37101 and PRJEB37035]; data supplement on the GSA Figshare Portal.

microbiology↗

bioRxiv2020

Efficacy of transoral laser surgery versus linear accelerator radiotherapy for the treatment of T1 glottic carcinoma: a meta-analysis of oncologic outcomes

ObjectiveA meta-analysis was conducted to compare oncologic outcomes for patients of T1 glottic carcinoma who were treated with transoral laser surgery (TLS) or linear accelerator radiotherapy (linac RT). MethodsAll related studies published up to September 2019 were acquired by searching Pubmed, EMBASE, and Cochrane, with the index words: glottic, vocal, laryngeal, radiation, radiotherapy, irradiation, laser, surgery, cordectomy, carcinoma and cancer. Relative studies which compared oncologic outcomes between linac RT and TLS were included. Sensitivity analysis were performed to evaluate heterogeneity. ResultsA total of twelve eligible studies were included for the analysis, which contained three prospective studies and nine retrospective studies. Patients who underwent TLS had increased overall survival (OR = 1.40, 95% CI=1.02-1.94, P=0.04) and laryngeal preservation (OR = 5.37, 95% CI = 3.05-9.44, P < 0.00001) versus who underwent linac RT. No statistical difference was observed between TLS group and linac RT group in terms of local control (OR=0.88, 95% CI = 0.62-1.24, P = 0.47), disease-specific survival (OR = 0.61, 95% CI = 0.26-1.43, P = 0.26), and disease-free survival (OR = 1.63, 95% CI = 0.70-3.81, P = 0.14). ConclusionsThe results of this meta-analysis indicate that there were clinical benefits for patients with glottic carcinoma after TLS compared with linac RT with respect to overall survival and laryngeal preservation. However, more multi-center randomized controlled trials would be urgently needed to prove these differences.

cancer biology↗

bioRxiv2020

Glycosaminoglycan binding motif at S1/S2 proteolytic cleavage site on spike glycoprotein may facilitate novel coronavirus (SARS-CoV-2) host cell entry

Severe acute respiratory syndrome-related coronavirus 2 (SARS-CoV-2) has resulted in a pandemic and continues to spread around the globe at an unprecedented rate. To date, no effective therapeutic is available to fight its associated disease, COVID-19. Our discovery of a novel insertion of glycosaminoglycan (GAG)-binding motif at S1/S2 proteolytic cleavage site (681-686 (PRRARS)) and two other GAG-binding-like motifs within SARS-CoV-2 spike glycoprotein (SGP) led us to hypothesize that host cell surface GAGs might be involved in host cell entry of SARS-CoV-2. Using a surface plasmon resonance direct binding assay, we found that both monomeric and trimeric SARS-CoV-2 spike more tightly bind to immobilized heparin (KD = 40 pM and 73 pM, respectively) than the SARS-CoV and MERS-CoV SGPs (500 nM and 1 nM, respectively). In competitive binding studies, the IC50 of heparin, tri-sulfated non-anticoagulant heparan sulfate, and non-anticoagulant low molecular weight heparin against SARS-CoV-2 SGP binding to immobilized heparin were 0.056 M, 0.12 M, and 26.4 M, respectively. Finally, unbiased computational ligand docking indicates that heparan sulfate interacts with the GAG-binding motif at the S1/S2 site on each monomer interface in the trimeric SARS-CoV-2 SGP, and at another site (453-459 (YRLFRKS)) when the receptor-binding domain is in an open conformation. Our study augments our knowledge in SARS-CoV-2 pathogenesis and advances carbohydrate-based COVID-19 therapeutic development.

biochemistry↗

bioRxiv2020

Optimization of multiplexed CRISPR/Cas9 system for highly efficient genome editing in Setaria viridis

In recent years, Setaria viridis has been developed as a model plant to better understand the C4 photosynthetic pathway in major crops. With the increasing availability of genomic resources for S. viridis research, highly efficient genome editing technologies are needed to create genetic variation resources for functional genomics. Here, we developed a protoplast assay to rapidly optimize the multiplexed CRISPR/Cas9 system in S. viridis. Targeted mutagenesis efficiency was further improved by an average of 1.4-fold with the exonuclease, Trex2. Distinctive mutation profiles were found in the Cas9_Trex2 samples with 94% of deletions larger than 10bp, and less than 1% of mutations being insertions. Further analyses indicated that 52.2% of deletions induced by Cas9_Trex2, as opposed to 3.5% by Cas9 alone, were repaired through microhomology-mediated end joining (MMEJ) rather than the canonical NHEJ DNA repair pathway. Combined with the robust agrobacterium-mediated transformation method with more than 90% efficiency, the multiplex CRISPR/Cas9_Trex2 system was demonstrated to induce targeted mutations in two tightly linked genes, svDrm1a and svDrm1b, at the frequency ranging from 73% to 100% in T0 plants. These mutations were transmitted to at least 60% of the transgene-free T1 plants with 33% of them containing bi-allelic or homozygous mutations in both genes. This highly efficient multiplex CRISPR/Cas9_Trex2 system makes it possible to create a large mutant resource for S. viridis in a rapid and high throughput manner, and has the potential to be widely applicable in achieving more predictable MMEJ-mediated mutations in many plant species.

plant biology↗

bioRxiv2020

LAMP-Seq: Population-Scale COVID-19 Diagnostics Using a Compressed Barcode Space

The ongoing SARS-CoV-2 pandemic has already caused devastating losses. Exponential spread can be slowed by social distancing and population-wide isolation measures, but those place a tremendous burden on society, and, once lifted, exponential spread can re-emerge. Regular population-scale testing, combined with contact tracing and case isolation, should help break the cycle of transmission, but current detection strategies are not capable of such large-scale processing. Here we present a protocol for LAMP-Seq, a barcoded Reverse-Transcription Loop-mediated Isothermal Amplification (RT-LAMP) method that is highly scalable. Individual samples are stabilized, inactivated, and amplified in three isothermal heat steps, generating barcoded amplicons that can be pooled and analyzed en masse by sequencing. Using unique barcode combinations per sample from a compressed barcode space enables extensive pooling, potentially further reducing cost and simplifying logistics. We validated LAMP-Seq on 28 clinical samples, empirically optimized the protocol and barcode design, and performed initial safety evaluation. Relying on world-wide infrastructure for next-generation sequencing, and in the context of population-wide sample collection, LAMP-Seq could be scaled to analyze millions of samples per day.

molecular biology↗

bioRxiv2020

Inhibition of nonsense-mediated decay rescues functional p53β/γ isoforms in MDM2-amplified cancers

Common mechanisms for p53 loss in cancer include expression of MDM2 or the human papilloma virus (HPV)-encoded E6 protein which both mediate degradation of wild-type (WT) p53 (p53). Here, we show that two alternatively-spliced, functional, truncated isoforms of p53 (p53{beta} and p53{gamma}, containing exons 1-9 of the p53 gene) can be markedly upregulated by pharmacologic or genetic inhibition of nonsense mediated decay (NMD), a regulator of aberrant mRNA stability. These isoforms lack the MDM2 binding domain and hence have reduced susceptibility to MDM2-mediated degradation. In MDM2-overexpressing cells bearing wildtype TP53 gene, NMD blockade increased p53{beta}/{gamma} expression and p53 pathway activation, enhanced radiosensitivity, and inhibited tumor growth. A similar pattern was observed in HPV+ cancer cells and in cancer cells with p53 mutations downstream of exon 9. These results identify a novel therapeutic strategy for restoration of p53 function in tumors rendered p53 deficient through MDM2 overexpression, HPV infection, or certain p53 mutations.

cancer biology↗

bioRxiv2020

WOX9 functions antagonistic to STF and LAM1 to regulate leaf blade expansion in Medicago truncatula and Nicotiana sylvestris

Plant specific WOX family transcription factors are known to regulate embryogenesis, meristem maintenance and lateral organ development. Modern clade WOX genes function through a transcriptional repression mechanism, and the intermediate clade transcriptional activator WOX9 functions with the repressor WOX genes in embryogenesis and meristems maintenance, but the mechanism of this interaction is unclear. WOX1 homologues STF and LAM1 are required for leaf blade outgrowth in Medicago truncatula and Nicotiana Sylvestris, respectively. Here we show that WOX9 negatively regulates leaf blade outgrowth and functions antagonistically to STF and LAM1. While NsWOX9 ectopic expression enhances the lam1 mutant phenotype, and antisense expression partially rescues the lam1 mutant, both overexpression of NsWOX9 and knockout by CRISPR/Cas9 genome editing in N. sylvestris resulted in a range of severe leaf blade distortions, indicating that controlled negative regulation by NsWOX9 is required for proper blade development. Our results indicate that direct repression of WOX9 transcriptional activation activity by the transcriptional repressor STF/LAM1 is required for correct blade architecture and patterning in M. truncatula and N. sylvestris. These findings suggest that a balance between transcriptional activation and repression mechanisms by direct interaction of activator and repressor WOX genes may be required for cell proliferation and differentiation homeostasis, and could be an evolutionarily conserved mechanism for the development of complex and diverse morphology in higher plants. One sentence summaryWOX9 negatively regulates blade outgrowth antagonizing STF function but directly repressed by STF indicating WOX-mediated homeostasis in cell proliferation and differentiation during leaf morphogenesis.

plant biology↗

bioRxiv2020

BatchServer: a web server for batch effect evaluation, visualization and correction

Batch effects are unwanted data variations that may obscure biological signals, leading to bias or errors in subsequent data analyses. Effective evaluation and elimination of batch effects are necessary for omics data analysis. In order to facilitate the evaluation and correction of batch effects, here we present BatchSever, an open-source R/Shiny based user-friendly interactive graphical web platform for batch effects analysis. In BatchServer we introduced autoComBat, a modified version of ComBat, which is the most widely adopted tool for batch effect correction. BatchServer uses PVCA (Principal Variance Component Analysis) and UMAP (Manifold Approximation and Projection) for evaluation and visualizion of batch effects. We demonstate its application in multiple proteomics and transcriptomic data sets. BatchServer is provided at https://lifeinfo.shinyapps.io/batchserver/ as a web server. The source codes are freely available at https://github.com/guomics-lab/batch_server.

bioinformatics↗

bioRxiv2020

Integrated analysis of a compendium of RNA-Seq datasets for splicing factors

A vast amount of public RNA-sequencing datasets have been generated and used widely to study transcriptome mechanisms. These data offer precious opportunity for advancing biological research in transcriptome studies such as alternative splicing. We report the first large-scale integrated analysis of RNA-Seq data of splicing factors for systematically identifying key factors in diseases and biological processes. We analyzed 1,321 RNA-Seq libraries of various mouse tissues and cell lines, comprising more than 6.6 TB sequences from 75 independent studies that experimentally manipulated 56 splicing factors. Using these data, RNA splicing signatures and gene expression signatures were computed, and signature comparison analysis identified a list of key splicing factors in Rett syndrome and cold-induced thermogenesis. We show that cold-induced RNA-binding proteins rescue the neurite outgrowth defects in Rett syndrome using neuronal morphology analysis, and we also reveal that SRSF1 and PTBP1 are required for energy expenditure in adipocytes using metabolic flux analysis. Our study provides an integrated analysis for identifying key factors in diseases and biological processes and highlights the importance of public data resources for identifying hypotheses for experimental testing.

bioinformatics↗

bioRxiv2020

A novel HER2-targeted antibody-drug conjugate offers the possibility of clinical dosing at trastuzumab-equivalent exposure levels

Trastuzumab and the related antibody-drug conjugate (ADC), ado-trastuzumab emtansine (T-DM1), both target HER2-overexpressing cells. Together, these drugs have treatment indications in both early-stage and metastatic settings for HER2+ breast cancer. T-DM1 retains the antibody functionalities of trastuzumab and adds the potency of a cytotoxic maytansine payload. Interestingly, in the clinic, T-DM1 cannot always replace the use of trastuzumab plus chemotherapy administered together as single agents. We hypothesize that this failure may be due in part to the limited systemic exposure achieved by T-DM1 relative to trastuzumab because of toxicity-related dosing constraints on the ADC. We have developed a trastuzumab-based ADC site-specifically conjugated to maytansine through a noncleavable linker. This construct, termed CAT-01-106, has a drug-to-antibody ratio (DAR) of 1.8, approximately half the average DAR of T-DM1, which comprises a mixture of antibodies variously conjugated with DARs ranging from 0-8. The high DAR species present in T-DM1 contribute to its toxicity and limit its clinical dose. CAT-01-106 showed superior in vivo efficacy compared to T-DM1 at equal payload dosing and was equally or better tolerated compared to T-DM1 at equal payload dosing up to 120 mg/kg in Sprague-Dawley rats and 60 mg/kg in cynomolgus monkeys. CAT-01-106 also showed improved pharmacokinetics in rats relative to T-DM1, with 40% higher ADC exposure levels. Together, the data suggest that CAT-01-106 may be sufficiently tolerable to enable clinical dosing at trastuzumab-equivalent exposure levels, combining the functions of both the antibody and the payload in one drug and potentially improving patient outcomes.

pharmacology and toxicology↗

bioRxiv2020

Filtering Variables for Supervised Sparse Network Analysis

MotivationWe present a method for dimension reduction designed to filter variables or features such as genes considered to be irrelevant for a downstream analysis designed to detect supervised gene networks in sparse settings. This approach can improve interpret-ability for a variety of analysis methods. We present a method to filter genes and transcripts prior to network analysis. This method has applications in a setting where the downstream analysis may include sparse canonical correlation analysis. ResultsFiltering methods specifically for cluster and network analysis are introduced and compared by simulating modular networks with known statistical properties. Our proposed method performs favorably eliminating irrelevant features but maintaining important biological signal under a variety of different signal settings. We show that the speed and accuracy of methods such as sparse canonical correlation are increased after filtering, thus greatly improving the scalability of these approaches. AvailabilityCode for performing the gene filtering algorithm described in this manuscript may be accessed through the geneFiltering R package available on Github at https://github.com/lorinmil/geneFiltering. Functions are available to filter genes and perform simulations of a network system. For access to the data used in this manuscript, contact corresponding author. Contactlorinmil@buffalo.edu, jcm38@buffalo.edu, fzhang8@buffalo.edu, and dlt6@buffalo.edu

bioinformatics↗

bioRxiv2020

Mapping the Structural and Dynamic Determinants of pH-sensitive Heparin Binding to Granulocyte Macrophage-colony Stimulating Factor

GMCSF is an immunomodulatory cytokine that is harnessed as a therapeutic. GMCSF is known to interact with other clinically important molecules, such as heparin, suggesting that endogenous and administered GMCSF has the potential to modulate orthogonal treatment outcomes. Thus, molecular level characterization of GMCSF and its interactions with other biologically active compounds is critical to understanding these mechanisms and predicting clinical outcomes. Here, we dissect the molecular motions and structural contributions that facilitate the GMCSF-heparin interaction, previously shown to be pH-dependent, using NMR spectroscopy, SPR, and molecular docking. We find that GMCSF and heparin binding is related to a change in flexibility reflected in the dynamic profile of GMCSF at acidic pH. The molecular motions driving this interaction largely occur on the ms-{micro}s timescale. Interestingly, we find that GMCSF and heparin binding is not only pH-dependent but is also heparin chain length dependent. We propose a mechanism where a positive binding pocket that is not fully solvent accessible at neutral pH becomes more accessible at acidic pH, allowing heparin to dock with the protein.

biochemistry↗

bioRxiv2020

Phenotype Prediction using a Tensor Representation and Deep Learning from Data Independent Acquisition Mass Spectrometry

A novel approach for phenotype prediction is developed for mass spectrometric data. First, the data-independent acquisition (DIA) mass spectrometric data is converted into a novel file format called "DIA tensor" (DIAT) which contains all the peptide precursors and fragments information and can be used for convenient DIA visualization. The DIAT format is fed directly into a deep neural network to predict phenotypes without the need to identify peptides or proteins. We applied this strategy to a collection of 102 hepatocellular carcinoma samples and achieved an accuracy of 96.8% in classifying malignant from benign samples. We further applied refined model to 492 samples of thyroid nodules to predict thyroid cancer; and achieved a predictive accuracy of 91.7% in an independent cohort of 216 test samples. In conclusion, DIA tensor enables facile 2D visualization of DIA proteomics data as well as being a new approach for phenotype prediction directly from DIA-MS data.

bioinformatics↗