Biology subjects

Lee, W.

Publications and source records attributed to Lee, W..

At least 73 records · Page 4Linked to original sources

bioRxiv2021

Natural Killer cells demonstrate distinct eQTL and transcriptome-wide disease associations, highlighting their role in autoimmunity

Natural Killer (NK) cells are innate lymphocytes with central roles in immunosurveillance and are implicated in autoimmune pathogenesis. The degree to which regulatory variants affect NK gene expression is poorly understood. We performed expression quantitative trait locus (eQTL) mapping of negatively selected NK cells from a population of healthy Europeans (n=245). We find a significant subset of genes demonstrate eQTL specific to NK cells and these are highly informative of human disease, in particular autoimmunity. An NK cell transcriptome-wide association study (TWAS) across five common autoimmune diseases identified further novel associations at 27 genes. In addition to these cis observations, we find novel master-regulatory regions impacting expression of trans gene networks at regions including 19q13.4, the Killer cell Immunoglobulin-like Receptor (KIR) Region, GNLY and MC1R. Our findings provide new insights into the unique biology of NK cells, demonstrating markedly different eQTL from other immune cells, with implications for disease mechanisms.

genetics↗

bioRxiv2021

Chitinase 3-like-1 Contributes to Acetaminophen-induced Liver Injury by Promoting Hepatic Platelet Recruitment

Hepatic platelet accumulation contributes to acetaminophen (APAP)-induced liver injury (AILI). However, little is known about the molecular pathways involved in platelet recruitment to the liver and whether targeting such pathways could attenuate AILI. The present study unveiled a critical role of chitinase 3-like-1 (Chi3l1) in hepatic platelet recruitment during AILI. Increased Chi3l1 and platelets in the liver were observed in patients and mice overdosed with APAP. Compared to wild-type (WT) mice, Chi3l1-/- mice developed attenuated AILI with markedly reduced hepatic platelet accumulation. Mechanistic studies revealed that Chi3l1 signaled through CD44 on macrophages to induce podoplanin expression, which mediated platelet recruitment through C-type lectin-like receptor 2. Moreover, APAP treatment of CD44-/- mice resulted in much lower numbers of hepatic platelets and liver injury than WT mice, a phenotype similar to that in Chi3l1-/- mice. Recombinant Chi3l1 could restore hepatic platelet accumulation and AILI in Chi3l1-/- mice, but not in CD44-/- mice. Importantly, we generated anti-Chi3l1 monoclonal antibodies and demonstrated that they could effectively inhibit hepatic platelet accumulation and AILI. Overall, we uncovered the Chi3l1/CD44 axis as a critical pathway mediating APAP-induced hepatic platelet recruitment and tissue injury. We demonstrated the feasibility and potential of targeting Chi3l1 to treat AILI.

pharmacology and toxicology↗

bioRxiv2021

Expression Based Species Deconvolution and Realignment Removes Misalignent Error in Multi-species Single Cell Data

Although single-cell RNA sequencing of xenograft samples has been widely used, no comprehensive bioinformatics pipeline is available for human and mouse mixed single-cell analyses. Considering the numerous homologous genes across the human and mouse genomes, misalignment errors should be evaluated, and a new algorithm is required. We assessed the extents and effects of misalignment errors when using human and mouse combined reference data and developed a new bioinformatics pipeline with expression-based species deconvolution to minimize errors. We also evaluated false-positive signals for a species presumed to originate from ambient RNA of the other species and applied a computational method to remove them. Misaligned reads account for an average of 0.5% of total reads, but such reads were concentrated to few genes that were greatly affected. Human and mouse mixed single-cell data, analyzed using our pipeline, clustered well with unmixed data and showed higher k-nearest-neighbor batch effect test and Local Inverse Simpsons Index scores than those derived from Cell Ranger. We also applied our pipeline to multispecies multisample single-cell library containing breast cancer xenograft tissue and successfully identified all samples using genomic array and expression. Moreover, diverse cell types in the tumor microenvironment were well captured. We present our bioinformatics pipeline for mixed human and mouse single-cell data, which can also be applied to pooled libraries to obtain cost-effective single-cell data. We also address major consideration points when analyzing multispecies single-cell data for other applications.

bioinformatics↗

bioRxiv2021

CCR2 Regulates Vaccine-Induced Mucosal T cell Memory to Influenza A Virus

Elicitation of lung tissue-resident memory CD8 T cells (TRMs) is a goal of T-cell based vaccines against respiratory viral pathogens such as influenza A virus (IAV). Chemokine receptor 2 (CCR2)-dependent monocyte trafficking plays an essential role in the establishment of CD8 TRMs in lungs of IAV-infected mice. Here, we used a combination adjuvant-based subunit vaccine strategy that evokes multifaceted (TC1/TC17/TH1/TH17) IAV nucleoprotein-specific lung TRMs, to determine whether CCR2 and monocyte infiltration are essential for vaccine-induced TRM development and protective immunity to IAV in lungs. Following intranasal vaccination, neutrophils, monocytes, conventional dendrtitic cells (DCs) and monocyte-derived DCs internalized and processed vaccine antigen in lungs. We also found that Basic Leucine Zipper ATF-Like Transcription Factor 3 (BATF-3)-dependent DCs were essential for eliciting T cell responses, but CCR2 deficiency enhanced the differentiation of CD127HI/KLRG-1LO, OX40+veCD62L+ve and mucosally imprinted CD69+veCD103+ve effector and memory CD8 T cells in lungs and airways of vaccinated mice. Mechanistically, increased development of lung TRMs, induced by CCR2 deficiency was linked to dampened expression of T-bet, but not altered TCF-1 levels or T cell receptor signaling in CD8 T cells. T1/T17 functional programming, parenchymal localization of CD8/CD4 effector and memory T cells, recall T cell responses and protective immunity to a lethal IAV infection were unaffected in CCR2-deficient mice. Taken together, we identified a negative regulatory role for CCR2 and monocyte trafficking in mucosal imprinting and differentiation of vaccine-induced TRMs. Mechanistic insights from this study may aid the development of T-cell-based vaccines against respiratory viral pathogens including IAV and SARS-CoV-2. ImportanceWhile antibody-based immunity to influenza A virus (IAV) is type and sub-type specific, lung and airway-resident memory T cells that recognize conserved epitopes in the internal viral proteins are known to provide heterosubtypic immunity. Hence, broadly protective IAV vaccines need to elicit robust T-cell memory in the respiratory tract. We have developed a combination adjuvant-based IAV nucleoprotein vaccine that elicits strong CD4 and CD8 T cell memory in lungs and protects against H1N1 and H5N1 strains of IAV. In this study, we examined the mechanisms that control vaccine-induced protective memory T cells in the respiratory tract. We found that trafficking of monocytes into lungs might limit the development of anti-viral lung-resident memory T cells, following intranasal vaccination. These findings suggested that strategies that limit monocyte infiltration can potentiate vaccine-induced frontline T-cell immunity to respiratory viruses such as IAV and SARS-CoV-2.

immunology↗

bioRxiv2020

Subunit Vaccine-Elicited Effector-Like Memory CD8 T cells Protect Against Listeriosis

Development of T-cell-based subunit protein vaccines against diseases, such as AIDS, tuberculosis and malaria remains a challenge for immunologists. Here, we have evaluated whether cross-presentation induced by nanoemulsion adjuvant Adjuplex (ADJ), can be combined with the immunomodulatory effects of TLR agonists (CpG or glucopyranosyl lipid adjuvant [GLA]) to evoke protective systemic CD8 T cell-based immunity to Listeria monocytogenes (LM). Vaccination with ADJ, alone or in combination with CpG or GLA augmented activation and antigen uptake by migratory and resident dendritic cells and up-regulated CD69 expression on B and T lymphocytes in draining lymph nodes. By virtue of its ability to engage BATF3-dependent cross-presenting DCs, ADJ potently elicited effector CD8 T cells that differentiated into a distinct subset of granzyme B-expressing CD27LO effector-like memory CD8 T cells, which provided highly effective immunity to LM in spleen and liver. CpG or GLA alone did not elicit effector-like memory CD8 T cells and induced moderate protection in spleen, but not in the liver. Surprisingly, combining CpG or GLA with ADJ limited the magnitude of ADJ-induced CD8 T cell memory and compromised protective immunity to LM, especially in the liver. Taken together, data presented in this manuscript provides a glimpse of protective CD8 T cell memory differentiation induced by a nano-emulsion adjuvant and demonstrates the unexpected negative effects of TLR signaling on the magnitude of CD8 T cell memory and protective immunity to listeriosis. ImportanceTo date, the most effective vaccines primarily provide protection by eliciting neutralizing antibodies, while development of T-cell-based subunit vaccines against infectious diseases, such as tuberculosis and malaria, remains a challenge for immunologists. Axiomatically, engagement of multiple innate immune receptors early in the response might be key to programming effective immunity. Hence, there is an impetus to develop combination adjuvants that engage multiple innate signaling pathways and additionally promote cross-presentation to stimulate CD8 T-cell immunity. Here, we show that a nano-emulsion adjuvant ADJ alone elicits effector-like memory CD8 T cells and provides highly effective immunity to listeriosis; combining ADJ with TLR agonists, including CpG and GLA, compromised T cell immunity to LM. In summary, this study provided fundamental insights into the effects of combining innate immune signaling with nano-emulsion adjuvants on memory T cell differentiation and protective immunity. These findings are expected to have implications in the use of combination adjuvants to develop subunit vaccines that engender systemic CD8 T-cell immunity to intracellular pathogens.

immunology↗

bioRxiv2020

Targeted long-read sequencing resolves complex structural variants and identifies missing disease-causing variants

BACKGROUNDDespite widespread availability of clinical genetic testing, many individuals with suspected genetic conditions do not have a precise diagnosis. This limits their opportunity to take advantage of state-of-the-art treatments. In such instances, testing sometimes reveals difficult-to-evaluate complex structural differences, candidate variants that do not fully explain the phenotype, single pathogenic variants in recessive disorders, or no variants in specific genes of interest. Thus, there is a need for better tools to identify a precise genetic diagnosis in individuals when conventional testing approaches have been exhausted. METHODSTargeted long-read sequencing (T-LRS) was performed on 33 individuals using Read Until on the Oxford Nanopore platform. This method allowed us to computationally target up to 100 Mbp of sequence per experiment, resulting in an average of 20x coverage of target regions, a 500% increase over background. We analyzed patient DNA for pathogenic substitutions, structural variants, and methylation differences using a single data source. RESULTSThe effectiveness of T-LRS was validated by detecting all genomic aberrations, including single-nucleotide variants, copy number changes, repeat expansions, and methylation differences, previously identified by prior clinical testing. In 6/7 individuals who had complex structural rearrangements, T-LRS enabled more precise resolution of the mutation, which led, in one case, to a change in clinical management. In nine individuals with suspected Mendelian conditions who lacked a precise genetic diagnosis, T-LRS identified pathogenic or likely pathogenic variants in five and variants of uncertain significance in two others. CONCLUSIONST-LRS can accurately predict pathogenic copy number variants and triplet repeat expansions, resolve complex rearrangements, and identify single-nucleotide variants not detected by other technologies, including short-read sequencing. T-LRS represents an efficient and cost-effective strategy to evaluate high-priority candidate genes and regions or to further evaluate complex clinical testing results. The application of T-LRS will likely increase the diagnostic rate of rare disorders.

genomics↗

bioRxiv2020

Effect of relative social rank within a social hierarchy on neural activation in response to familiar or unfamiliar social signals

Competent social functioning of group-living species relies on the ability of individuals to detect and utilize conspecific social cues to guide behavior. Previous studies have identified numerous brain regions involved in processing these external cues, collectively referred to as the Social Decision-Making Network. However, how the brain encodes social information with respect to an individuals social status has not been thoroughly examined. In mice, cues about an individuals identity, including social status, are conveyed through urinary proteins. In this study, we assessed the neural cFos immunoreactivity in dominant and subordinate male mice exposed to familiar and unfamiliar dominant and subordinate male urine. The posteroventral medial amygdala was the only brain region that responded exclusively to dominant compared to subordinate male urine. In all other brain regions, including the VMH, PMv, and vlPAG, activity is modulated by a combination of odor familiarity and the social status of both the urine donor and the subject receiving the cue. We show that dominant subjects exhibit robust differential activity across different types of cues compared to subordinate subjects, suggesting that individuals perceive social cues differently depending on social experience. These data inform further investigation of neurobiological mechanisms underlying social-status related brain differences and behavior.

neuroscience↗

bioRxiv2020

Solution structures of the Shewanella woodyi H-NOX protein in the presence and absence of soluble guanylyl cyclase stimulator IWP-051

Heme-nitric oxide/oxygen binding (H-NOX) domains bind gaseous ligands for signal transduction in organisms spanning prokaryotic and eukaryotic kingdoms. In the bioluminescent marine bacterium Shewanella woodyi (Sw), H-NOX proteins regulate quorum sensing and biofilm formation. In higher animals, soluble guanylyl cyclase (sGC) binds nitric oxide with an H-NOX domain to induce cyclase activity and regulate vascular tone, wound healing and memory formation. sGC also binds stimulator compounds targeting cardiovascular disease. The molecular details of stimulator binding to sGC remain obscure but involve a binding pocket near an interface between H-NOX and coiled-coil domains. Here, we report the full NMR structure for CO-ligated Sw H-NOX in the presence and absence of stimulator compound IWP-051, and its backbone dynamics. Non-planar heme geometry was retained using a semi-empirical quantum potential energy approach. Although IWP-051 binding is weak, a single binding conformation was found at the interface of the two H-NOX subdomains. Binding lead to rotation of the subdomains and closure of the binding pocket. Backbone dynamics for the protein are similar across both domains except for two helix-connecting loops, which display increased dynamics that are further enhanced by compound binding. Structure-based sequence analyses indicate high sequence diversity in the binding pocket, but the pocket itself appears conserved among H-NOX proteins. The largest dynamical loop lies at the interface between Sw H-NOX and its binding partner as well as in the interface with the coiled coil in sGC, suggesting a critical role for the loop in signal transduction.

biochemistry↗

bioRxiv2020

Osteoarthritis: articular chondrocyte inflammatory signaling leads to enhanced gene expression and function of mechanotransduction channel Piezo1 as a pathogenic feedforwardmechanism

Osteoarthritis (OA) is a painful and debilitating condition of synovial joints without any disease-modifying therapies (1, 2). We previously identified mechanosensitive PIEZO channels, PIEZO1 and PIEZO2, both expressed in articular cartilage, to function in chondrocyte mechanotransduction in response to injury (3, 4). We therefore asked whether interleukin-1-mediated inflammatory signaling, as occurs in OA, influences Piezo gene expression and channel function, thus indicative of maladaptive reprogramming that can be rationally targeted. Primary porcine chondrocyte culture and human osteoarthritic cartilage tissue were studied. We found that interleukin-1 (IL-1) upregulated Piezo1 in porcine chondrocytes. Piezo1 expression was significantly increased in human osteoarthritic cartilage. Increased Piezo1 expression in chondrocytes resulted in a feed-forward pathomechanism whereby increased function of Piezo1 induced excess intracellular Ca2+, at baseline and in response to mechanical deformation. Elevated resting state Ca2+ in turn rarefied the F-actin cytoskeleton and amplified mechanically-induced deformation-microtrauma. As intracellular substrates of this novel OA-related inflammatory pathomechanism, in porcine articular chondrocytes exposed to IL-1 we discovered that enhanced Piezo1 expression depended on p38 MAP-kinase and transcription factors HNF4 and ATF2/CREBP1. CREBP1 directly bound to the proximal PIEZO1 gene promoter. In ensemble, these signaling and genetic re-programming events represent a novel and detrimental Ca2+-driven feed-forward mechanism that can be rationally targeted to stem the progression of OA. Significance StatementOsteoarthritis affecting weight-bearing joints is a global health problem, causing loss of mobility and enormous healthcare costs. Disease-modifying approaches are lacking. Here, we report a new cellular mechanism of inflammatory signaling in chondrocytes, the cellular substrate of cartilage. We show how osteoarthritis-relevant levels of interleukin-1 reprogram articular chondrocytes so that they become more susceptible to mechanical trauma, which chondrocytes sense via Piezo1/2 mechanosensitive ion channels. We uncover that IL-1 enhances gene expression of Piezo1 in primary articular chondrocytes underlying Piezo1 gain-of-function. We elucidate the new signaling pathway, from membrane to nucleus, including transcription factors that enhance Piezo1-expression. We also define detrimental effects of gain-of-function of Piezo1, for mechanotransduction and at-rest, that suggest this new reprogramming mechanism to contribute to osteoarthritis pathogenesis.

molecular biology↗

bioRxiv2020

Nano-Emulsion Adjuvant Enhances Cross-presentation by Lipid Body Formation Independent of Glycolysis

Here, we report that a carbomer-based adjuvant, Adjuplex(R) (ADJ), stimulated robust CD8 T-cell responses to subunit antigens by modulating multiple steps in the cytosolic pathway of cross-presentation, and afforded effective immunity against virus and intracellular bacteria. Cross-presentation induced by TLR agonists requires a critical switch to anabolic metabolism, but ADJ enhanced cross presentation without this metabolic switch in DCs and NLRP3-driven caspase 1 activity. Instead, ADJ induced in DCs, an unique metabolic state, typified by dampened oxidative phosphorylation and basal levels of glycolysis. In the absence of increased glycolytic flux, induction of ROS and lipid bodies (LBs) and alterations in LB composition mediated by ADJ were critical for DC cross-presentation. These findings challenge the prevailing metabolic paradigm by suggesting that DCs can perform effective DC cross-presentation, independent of glycolysis to induce robust T cell-dependent protective immunity to intracellular pathogens. These findings have implications in the rational development of novel adjuvants.

immunology↗

bioRxiv2020

IER5, a DNA-damage response gene, is required for Notch-mediated induction of squamous cell differentiation

Notch signaling regulates normal squamous cell proliferation and differentiation and is frequently disrupted in squamous cell carcinomas, in which Notch is a key tumor suppressive pathway. To identify the direct targets of Notch that produce these phenotypes, we introduced a conditional Notch transgene into squamous cell carcinoma cell lines, which respond to Notch activation in 2D culture and in organoid cultures by undergoing differentiation. RNA-seq and ChIP-seq analyses show that in squamous cells Notch activates a context-specific program of gene expression that depends on lineage-specific regulatory elements, most of which lie in long- range enhancers. Among the direct Notch target genes are multiple DNA damage response genes, including IER5, which is regulated by Notch through several enhancer elements. We show that IER5 is required for Notch-induced differentiation in squamous carcinoma cells and in TERT-immortalized keratinocytes. Its function is epistatic to PPP2R2A, which encodes the B55 subunit of PP2A, and IER5 interacts with B55 in cells and in purified systems. These results show that Notch and DNA-damage response pathways converge in squamous cells and that some components of these pathways promote differentiation, which may serve to eliminate DNA-damaged cells from the proliferative pool in squamous epithelia. Crosstalk involving Notch and PP2A may enable Notch signaling to be tuned and integrated with other pathways that regulate squamous differentiation. Our work also suggests that squamous cell carcinomas remain responsive to Notch signaling, providing a rationale for reactivation of Notch as a therapeutic strategy in these cancers. ImpactOur findings highlight context-specific crosstalk between Notch, DNA damage response genes, and PP2A, and provide a roadmap for understanding how Notch induces the differentiation of squamous cells.

cancer biology↗

bioRxiv2019

Prediction of 8-state protein secondary structures by 1D-Inception and BD-LSTM

Protein secondary structure is crucial to create an information bridge between the primary structure and the tertiary (3D) structure. Precise prediction of 8-state protein secondary structure (PSS) significantly utilized in the structural and functional analysis of proteins in bioinformatics. In this recent period, deep learning techniques have been applied in this research area and raise the Q8 accuracy remarkably. Nevertheless, from a theoretical standpoint, there still lots of room for improvement, specifically in 8-state (Q8) protein secondary structure prediction. In this paper, we presented two deep learning architecture, namely 1D-Inception and BD-LSTM, to improve the performance of 8-classes PSS prediction. The input of these two architectures is a carefully constructed feature matrix from the sequence features and profile features of the proteins. Firstly, 1D-Inception is a Deep convolutional neural network-based approach that was inspired by the InceptionV3 model and containing three inception modules. Secondly, BD-LSTM is a recurrent neural network model which including bidirectional LSTM layers. Our proposed 1D-Inception method achieved 76.65%, 71.18%, 76.86%, and 74.07% Q8 accuracy respectively on benchmark CullPdb6133, CB513, CASP10, and CASP11 datasets. Moreover, BD-LSTM acquired 74.71%, 69.49%, 74.07%, and 72.37% state-8 accuracy after evaluated on CullPdb6133, CB513, CASP10, and CASP11 datasets, respectively. Both these architectures enable the efficient processing of local and global interdependencies between amino acids to make an accurate prediction of each class is very beneficial in the deep neural network. To the best of our knowledge, experiment results of the 1D-Inception model demonstrate that it outperformed all the state-of-art methods on the benchmark CullPdb6133, CB513, and CASP10 datasets.

bioinformatics↗

bioRxiv2019

Mapping cellular-scale internal stiffness in 3D tissues with smart material hydrogel probes

Local stiffness plays a critical role in cell function, but measuring rigidity at cellular length scales in living 3D tissues presents considerable challenges. Here we present thermoresponsive, smart material microgels that can be dispersed or injected into tissues and optically assayed to measure internal tissue stiffness over several weeks. We first develop the material design principles to measure tissue stiffness across physiological ranges, with spatial resolutions approaching that of individual cells. Using the microfabricated sensors, we demonstrate that mapping internal stiffness profiles of live multicellular spheroids at high resolutions reveal distinct architectural patterns, that vary with subtle differences in spheroid aggregation method. Finally, we determine that small sites of unexpectedly high stiffness (> 250 kPa) develop in invasive breast cancer spheroids, and in in vivo mouse model tumors as the cancer progresses towards metastatic disease. These highly focal sites of increased intratumoral stiffness likely form via active cell mechanical behavior, and suggest new possibilities for how early mechanical cues that drive cancer cells towards invasion might arise within the evolving tumor microenvironment.

biophysics↗

bioRxiv2019

Selective constraints and pathogenicity of mitochondrial DNA variants inferred from a novel database of 196,554 unrelated individuals

High quality population allele frequencies of DNA variants can be used to discover new biology, and study rare disorders. Here, we created a public catalog of mitochondrial DNA variants based on a population of 195,983 individuals. We focused on 3 criteria: (i) the population is not enriched for mitochondrial disorders, or other clinical phenotypes, (ii) all genomes are sequenced and analyzed in the same clinical laboratory, and (iii) both homoplasmic and heteroplasmic variants are reported. We found that 47% of the mitochondrial genome was invariant in this population, including large stretches in the 2 rRNA genes. This information could be used to annotate the mitochondrial genome in future studies. We also showed how to use this resource for the interpretation of pathogenic variants for rare mitochondrial disorders. For example, 42% of variants previously reported to be pathogenic for Leber Hereditary Optic Neuropathy (LHON) should be reclassified.

genetics↗

bioRxiv2019

Human Reference Genome and a High Contiguity Ethnic Genome AK1

Studies have shown that the current human reference genome (GRCh38) might miss information for some populations, but \"exactly what we miss\" is still elusive due to the lower contiguity of non-reference genomes. We juxtaposed the GRCh38 with high contiguity genome assemblies, AK1, to show that [~]1.8% ([~]53.4 Mbp) of AK1 sequences missed in GRCh38 with [~]0.76% ([~]22.2 Mbp) of ectopic chromosomes. The unique AK1 sequences harbored [~]1,390 putative coding elements. We found that [~]5.3Mb ([~]0.2%) of the AK1 sequences aligned and recovered the \"unmapped\" reads of fourteen individuals (5 East-Asians, 4 Europeans, and 5 Africans) as a reference. The regions that \"unmapped\" reads aligned included 110 common (shared between [≥]2 individuals) and 38 globally ([≥]7 individuals) missing regions with 25 candidate coding elements. We verified that many of the common missing regions exist in multiple populations and chimpanzees DNA. Our study illuminates not only the discovery of missing information but the use of highly precise ethnic genomes in understanding human genetics.

genomics↗

bioRxiv2019

Atrous Convolution with transfer learning for Skin Lesions Classification

Skin cancer is a crucial public health issue and by far the most usual kind of cancer specifically in the region of North America. It is estimated that in 2019, only because of melanoma nearly 7,230 people will die, and 192,310 cases of malignant melanoma will be diagnosed. Nonetheless, nearly all types of skin lesions can be treatable if they can be diagnosed at an earlier stage. The accurate prediction of skin lesions is a critically challenging task even for vastly experienced clinicians and dermatologist due to a little distinction between surrounding skin and lesions, visual resemblance between melanoma and other skin lesions, fuddled lesion border, etc. A well-grounded automated computer-aided skin lesions detection system can help clinicians immensely to prognosis malignant skin lesion in the earliest possible time. From the past few years, the emergence of machine learning and deep learning in the medical imaging has produced several image-based classification systems in the medical field and these systems perform better than traditional image processing classification methods. In this paper, we proposed a popular deep learning technique namely atrous or, dilated convolution for skin lesions classification, which is known to have enhanced accuracy with the same amount of computational cost compared to traditional CNN. To implement atrous convolution we choose the transfer learning technique with several popular deep learning architectures such as VGG16, VGG19, MobileNet, and InceptionV3. To train, validate, and test our proposed models we utilize HAM10000 dataset which contains total 10015 dermoscopic images of seven different skin lesions (melanoma, melanocytic nevi, Basal cell carcinoma, Benign keratosis-like lesions, Dermatofibroma, Vascular lesions, and Actinic keratoses). Four of our proposed dilated convolutional frameworks show promising outcome on overall accuracy and per-class accuracy. For example, overall test accuracy achieved 87.42%, 85.02%, 88.22%, and 89.81% on dilated VGG16, dilated VGG19, dilated MobileNet, and dilated IncaptionV3 respectively. These dilated convolutional models outperformed existing networks in both overall accuracy and individual class accuracy. Among all the architectures dilated InceptionV3 shows superior classification accuracy and dilated MobileNet is also achieving almost impressive classification accuracy like dilated InceptionV3 with the lightest computational complexities than all other proposed model. Compared to previous works done on skin lesions classification we have experimented one of the most complicated open-source datasets with class imbalances and achieved better accuracy (dilated inceptionv3) than any known methods to the best of our knowledge.

cancer biology↗

bioRxiv2019

Temporal microstructure of dyadic social behavior during relationship formation in mice

Understanding the temporal dynamics of how unfamiliar animals establish dominant-subordinate relationships and learn to modify their behavior in response to their social partner in context-appropriate manners is critical in biomedical research concerning social competence. Here we observe and analyze the microstructure of social and non-social behaviors as 21 pairs of outbred CD-1 male mice (Mus Musculus) establish dominant-subordinate relationships during daily 20-minute interaction for five consecutive days. Using Kleinberg burst detection algorithm, we demonstrate aggressive and subordinate interactions occur in bursting patterns followed by quiescence period rather than in uniformly distributed across social interactions. Further, we identify three phases of dominant-subordinate relationship development (pre-, middle-, and post-resolution) by combining phi-coefficient and difference methods used to determine at which bursting event mice resolve dominant-subordinate relationships. Using First Order Markov Chains within individuals we show dominant and subordinate animals establish significantly different behavioral repertoire once they resolve the relationships. In both dominant and subordinate mice, the transitions between investigative and agonistic behavior states are not common. Lastly, we introduce Forward Spike Time Tiling Coefficient, the strength of association between the given behavior of one individual with the target behavior of the other individual within a specified time window. With this method, we describe the likelihood of a mouse responding to a behavior with another behavior differ in pre- and post-resolution phases. The data suggest that subordinate mice learn to exhibit subordinate behavior in response to dominant partners behaviors while dominant mice become less likely to show subordinate behaviors in response to their partners action. Overall, with the tool we present in this study, the data suggest CD-1 male mice are able to establish dominance relationships and modify their behaviors even to the same social cues under different social contexts competently.

animal behavior and cognition↗

bioRxiv2019

Comparative Tn-Seq reveals common daptomycin resistance determinants in Staphylococcus aureus despite strain-dependent differences in essentiality of shared cell envelope genes

Antibiotic-resistant Staphylococcus aureus remains a leading cause of antibiotic resistance-associated mortality in the United States. Given the reality of multi-drug resistant infections, it is imperative that we establish and maintain a pipeline of new compounds to replace or supplement our current antibiotics. A first step towards this goal is to prioritize targets by identifying the genes most consistently required for survival across the S. aureus phylogeny. Here we report the first direct comparison of gene essentiality across multiple strains of S. aureus via transposon sequencing. We show that mutant fitness varies by strain in key pathways, underscoring the importance of using more than one strain to differentiate between core and strain-dependent essential genes. Despite baseline differences in gene importance, several pathways, including the lipoteichoic acid pathway, become consistently essential under daptomycin exposure, suggesting core vulnerabilities that can be exploited to resensitize daptomycin-nonsusceptible isolates. We also demonstrate the merit of using transposons with outward-facing promoters capable of overexpressing nearby genes for identifying clinically-relevant gain-of-function resistance mechanisms. Together, the daptomycin vulnerabilities and resistance mechanisms support a mode of action with wide-ranging effects on the cell envelope and cell division. This work adds to a growing body of literature demonstrating the nuanced insights gained by comparing Tn-Seq results across multiple bacterial strains.\n\nAuthor summaryAntibiotic-resistant Staphylococcus aureus kills thousands of people every year in the United States alone. To stay ahead of the looming threat of multidrug-resistant infections, we must continue to develop new antibiotics and find ways of making our current repertoire of antibiotics more effective, including by finding pairs of compounds that perform best when administered together. In the age of next-generation sequencing, we can now use transposon sequencing to find potential targets for new antibiotics on a genome-wide scale, identified as either essential genes or genes that become essential in the presence of an antibiotic. In this work, we created a compendium of genes that are essential across a range of S. aureus strains, as well as those that are essential in the presence of the antibiotic daptomycin. The results will be a resource for researchers working to develop the next generation of antibiotic therapies.

microbiology↗