Biology subjects

Pu, L.

Publications and source records attributed to Pu, L..

At least 19 recordsLinked to original sources

bioRxiv2026

PlasChain: an algorithm for improving long plasmid reconstruction from metagenome assemblies

BackgroundPlasmids play a critical role in horizontal gene transfer and the spread of antibiotic resistance. However, recovering complete plasmid sequences from metagenomic samples remains highly challenging due to extensive repeat content, structural heterogeneity, and large variation in plasmid size. Existing methods typically identify plasmids from metagenome assemblies by exploiting coverage differences or detecting minimum-weight cycles in assembly graphs. While effective for dominant plasmids, these approaches often fail to recover low-abundance and long plasmids. ResultsHere we present PlasChain, a novel algorithm designed to improve plasmid assembly and identification from complex metagenomic data. Building upon the cycle-peeling strategy of SCAPP, PlasChain incorporates contig path information and a cycle-merging procedure to prevent long plasmids from being fragmented into multiple shorter cycles. In addition, PlasChain jointly leverages paired-end read alignments, sequence composition patterns, and coverage variation to filter out assembly artifacts and reduce false positives. We evaluated PlasChain against state-of-the-art plasmid assemblers, including SCAPP and metaplasmidSPAdes, using a diverse set of simulated and real metagenomic datasets. Across nearly all benchmarks, PlasChain demonstrates superior performance in recovering long plasmids while maintaining competitive accuracy in assembling short plasmids. Furthermore, analysis of real metagenomic samples shows that PlasChain is capable of assembling previously uncharacterized plasmids, including putative megaplasmids that are typically underrepresented in current plasmid databases. ConclusionsPlasChain is a novel graph-based plasmid assembler that improves the recovery of long plasmids from short-read metagenomic data. Evaluation on diverse simulated and real metagenomic datasets demonstrates that PlasChain consistently outperforms existing plasmid assemblers, especially for long plasmid reconstruction. These results highlight the potential of PlasChain to facilitate comprehensive characterization of plasmid diversity, antimicrobial resistance, and horizontal gene transfer in complex microbial communities. The source code and testing data of PlasChain are freely available at https://github.com/SDU-ACG-Lab/PlasChain. Supplementary InformationThis manuscript is accompanied by a supplementary file.

bioinformatics↗

bioRxiv2025

Seq2Karyotype (S2K): A Method for in-silico Karyotyping Using Single-Sample Whole-Genome Sequencing Data

DNA abnormalities characterized by cytogenetic imaging at the single cell resolution, i.e. karyotyping, have long served as cancer diagnostic and prognostic biomarkers. To enable in-silico karyotyping using unpaired whole-genome sequencing data, we developed Seq2Karyotype (S2K), a tool that fits karyotype models with clonality estimation based on read-depth and allelic imbalance in a bulk sample and supports visualization-guided refinement. Analysis on 19 adult and pediatric cancer cell lines revealed unexpected intratumoral heterogeneity involving multiple copy number variation (CNV) states including whole-genome duplication, which were validated by imaging and single-cell omics profiling. Analyses on patient samples showed high concordance with clinical cytogenetic reports for acute myeloid leukemia, and revealed evolutionary trajectories from multi-region metastatic neuroblastomas implicating reversion. These findings highlight extensive and dynamic intratumoral heterogeneity contributed by CNV in both cell line models and patient samples, which may inform future research on tumor evolution under selective pressure such as drug exposure.

bioinformatics↗

bioRxiv2025

The bivalent epigenetic regulator ULTRAPETALA1 promotes the Arabidopsis floral transition via recruitment of Polycomb histone methyltransferases for H3K27me3 deposition at the FLC locus

Floral induction is a major transition during the plant life cycle that contributes to reproductive fitness in annual and perennial plants. Flowering occurs in response to multiple environmental and endogenous cues. In Arabidopsis thaliana, many of these cues converge on the regulation of the FLOWERING LOCUS C (FLC) gene, which encodes a MADS domain transcription factor that functions as the main floral repressor. Here we demonstrate that the bivalent epigenetic regulator ULTRAPETALA1 (ULT1) is both necessary and sufficient to promote flowering. We identify ULT1 as a novel component of the autonomous pathway (AP) and show that it accelerates flowering by directly repressing FLC transcription. We demonstrate that ULT1 physically associates with the SWINGER (SWN) and CURLY LEAF (CLF) histone methyltransferase components of Polycomb Repressive Complex 2 (PRC2) to regulate FLC transcription by altering the accumulation of H3K27me3 marks. Our data indicate that ULT1 promotes the floral transition by interacting with SWN and CLF to recruit PRC2 to the FLC locus to deposit repressive histone methylation and reduce transcription of this key floral repressor.

plant biology↗

bioRxiv2025

Effects of Exercise and Konjac on Lipid Metabolism and Mechanism in Obese Rats

ObjectiveTo study the effects of exercise and konjac on body weight and lipid metabolism in obese rats, and the possible mechanism of exercise combined with konjac on body weight and lipid metabolism. MethodsFifty male Sprague Dawley rats were randomly divided into 5 groups. After six weeks of intervention, blood lipid, body weight, liver weight, inflammatory factors, oxidative stress markers and proteins were measured. ResultsWeight and liver weight in HFD group were significantly higher (P<0.05, P<0.01) than those in CON group. Compared with HFD group, blood lipid index in intervention group were significantly decreased; The levels of MDA (P<0.01, P<0.05, P<0.05) and SOD (P<0.05) in intervention group were significantly decreased; the GSH levels in HAE group and HKE group were significantly increased(P<0.01, P<0.05); the levels of IL-6, TNF-, and leptin in HAE group (P<0.05, P<0.05, P<0.01) and HKE group (P<0.01, P<0.01, P<0.05) were significantly decreased; ApoA5 in intervention group was significantly increased (P<0.01), SREBP-1c and LXR was significantly decreased (P<0.01). ConclusionKonjac or exercise alone can improve the lipid metabolism, and the comprehensive effect is better. The mechanism of regulating lipid metabolism may be related to the expression of LXR, SREBP-1c and ApoA5.

biochemistry↗

bioRxiv2025

Single-cell spatial transcriptome reveals pathological features of human hippocampus with sclerosis

More than half of epileptic patients ultimately turned to intractable epilepsy. Mesial temporal lobe epilepsy with hippocampal sclerosis (MTLE-HS), the most common type of intractable epilepsies, whose pathological mechanism remains elusive. Here, using 42 human hippocampal samples from surgical donors of MTLE (32 with and 10 without HS) through single-cell resolution Stereo-seq and histologcial experiments, we revealed spatially pathological changes of gene expression and cell type composition with HS systematically. After precise parcellation of hippocampal subregions and differentially expressed gene (DEG) analysis between each region with and without sclerosis, we found Cornu Ammonis (CA) subregions with higher number of DEGs were vulnerable to sclerosis, especially CA1 and CA3. Within CA1, we found that CA1-superficial and proximal areas were more vulnerable than CA1-deep and distal areas in sclerosis. Meanwhile, after analyzing 350,795 segmented cells from Stereo-seq, we found dramatically increasing density of astrocyte accompanied with significantly decreasing density of excitatory neurons in CA1, especially superficial and proximal CA1, and CA3 in sclerosis. In these two vulnerable subregions, proliferative astrocyte (P_astrocyte) and reactive astrocyte (R_astrocyte) were found to be enriched whereas apoptotic subtype of astrocyte (A_astrocyte), related with apoptotic pathway, was mainly located in alveus, which strengthened cell communication with reactive microglia (R_microglia) in HS, revealing the novel pathological feature in our work. The pseudotime analysis indicated that CA excitatory neurons underwent synaptic impairment, energy dysfunction, aging, and finally losing cell identity until death through autophagy or apoptosis. Besides, we also found a resilient subtype, EX_CA2-4.3, highly expressed extracellular matrix related genes including PDYN, and was increasing the interaction of BDNF-NTRK, NFASC-CNTN1 to withstand the damage from sclerosis. Together, our study provided a reference of human hippocampus with and without HS caused by MTLE, and highlighted the potential pathological mechanism on molecular and cellular level of MTLE-HS.

neuroscience↗

bioRxiv2024

Microglial states determine lesion dynamics in multiple sclerosis

Summary paragraphMultiple sclerosis (MS) is a neuroinflammatory disease of the central nervous system, characterized by demyelinating lesions 1. Lesion expansion contributes to progression and increased disability, while remyelination can recover neurological deficits. However, mechanisms driving lesion dynamics are largely unclear, hindering the development of effective therapeutics. We propose that distinct states of microglia are involved in lesion expansion and remyelination 2,3. Using Stereo-seq, an RNA capture based high-resolution spatial transcriptomics technology with single-cell resolution, on post-mortem human brain tissue, we compared mixed active/inactive lesions with lipid-laden foamy microglia with lesions containing ramified microglia. We identified distinct cellular and molecular mechanisms underlying lesion activity and remyelination, linked to microglia phenotypes and states. Lesions with foamy microglia were characterized by elevated immune activation, increased lymphocyte densities, upregulated immunoglobulin production (IGHG1, IGHG3), increased complement system activity, indication of iron dysregulation (FTL, FTH1), and increased demyelination. In contrast, lesions with ramified microglia exhibited gene expression profiles indicative of myelin stability (ABCA2, QKI) and neuro-axonal protection, fostering an environment conducive to repair and remyelination. Our findings highlight the role of microglial states in lesion expansion and repair in MS and offer promising avenues for the development of therapeutic approaches aimed at preventing MS disability progression.

neuroscience↗

bioRxiv2023

Acute avoidance of hydrogen sulfide is modulated by external and internal states in C. elegans

Hydrogen sulfide (H2S) acts as an energy source, a toxin, and a gasotransmitter across diverse biological contexts. We use the robust locomotory responses of Caenorhabditis elegans to high levels of H2S to elucidate the molecular mechanisms underlying its acute and adaptive responses. We find that the H2S-evoked behavioral response is shaped by multiple environmental factors including oxygen (O2) levels and nutritional state, and is modulated by various pathways such as insulin, TGF-{beta}, and HIF-1 signaling, as well as by input from O2-sensing neurons. Prolonged exposure to H2S activates HIF-1 signaling, leading to the upregulation of stress-responsive genes, including those involved in H2S detoxification. This promotes an adaptive state in which locomotory speed is reduced in H2S, while responsiveness to other stimuli is preserved. In mutants deficient in HIF-1 signaling, iron storage, and detoxification mechanisms, animals display a robust initial response but rapidly enter a sleep-like behavior characterized by reduced mobility and diminished responsiveness to subsequent sensory stimuli. Furthermore, while acute production of mitochondria-derived reactive O2 species (ROS) appears to initiate the avoidance response to H2S, persistently high ROS promotes an adaptive state, likely by activating various stress-response pathways, without substantially compromising cellular H2S detoxification capacity. Taken together, our study provides comprehensive molecular insights into the mechanisms through which C. elegans modulates and adapts its response to H2S exposure.

animal behavior and cognition↗

bioRxiv2023

4CAC: 4-class classification of metagenome assembliesusing machine learning and assembly graphs

Microbial communities usually harbor a mix of bacteria, archaea, plasmids, viruses, and microeukaryotes. Within these communities, viruses, plasmids, and microeukaryotes coexist in relatively low abundance, yet they engage in intricate interactions with bacteria. Moreover, viruses and plasmids, as mobile genetic elements, play important roles in horizontal gene transfer and the development of antibiotic resistance within microbial populations. However, due to the difficulty of identifying viruses, plasmids, and microeukaryotes in microbial communities, our understanding of these minor classes lags behind that of bacteria and archaea. Recently, several classifiers have been developed to separate one or two minor classes from bacteria and archaea in metagenome assemblies, but none can classify all of the four classes simultaneously. Moreover, existing classifiers have low precision on minor classes. Here, we developed a classifier called 4CAC that is able to identify viruses, plasmids, microeukaryotes, and prokaryotes simultaneously from metagenome assemblies. 4CAC generates an initial four-way classification using several sequence length-adjusted XGBoost models and further improves the classification using the assembly graph. Evaluation on simulated and real metagenome datasets demonstrates that 4CAC substantially outperforms existing classifiers and combinations thereof on short reads. On long reads, it also shows an advantage unless the abundance of the minor classes is very low. 4CAC runs 1-2 orders of magnitude faster than the other classifiers. The 4CAC software is available at https://github.com/Shamir-Lab/4CAC.

bioinformatics↗

bioRxiv2022

Efficient minimizer orders for large values of k using minimum decycling sets

Minimizers are ubiquitously used in data structures and algorithms for efficient searching, mapping, and indexing of high-throughput DNA sequencing data. Minimizer schemes select a minimum k-mer in every L-long sub-sequence of the target sequence, where minimality is with respect to a predefined k-mer order. Commonly used minimizer orders select more k-mers than necessary and therefore provide limited improvement in runtime and memory usage of downstream analysis tasks. The recently introduced universal k-mer hitting sets produce minimizer orders with fewer selected k-mers. Unfortunately, generating compact universal k-mer hitting sets is currently infeasible for k > 13, and thus cannot help in the many applications that require minimizer orders for larger k. Here, we close the gap of efficient minimizer orders for large values of k by introducing decycling-set-based minimizer orders, new orders based on minimum decycling sets. We show that in practice these new minimizer orders select a number of k-mers comparable to that of minimizer orders based on universal k-mer hitting sets, and can also scale up to larger k. Furthermore, we developed a method that computes the minimizers in a sequence on the fly without keeping the k-mers of a decycling set in memory. This enables the use of these minimizer orders for any value of k. We expect the new orders to improve the runtime and memory usage of algorithms and data structures in high-throughput DNA sequencing analysis.

bioinformatics↗

bioRxiv2022

Programming the lifestyles of engineered bacteria for cancer therapy

Bacteria can be genetically engineered to act as therapeutic delivery vehicles in the treatment of tumors, killing cancer cells or activating the immune system. This is known as Bacteria-Mediated Cancer Therapy (BMCT). Tumor invasion, colonization and tumor regression are major biological events, which are directly associated with antitumor effects and are uncontrollable due to the influence of tumor microenvironments during the BMCT process. Here, we developed a genetic circuit for dynamically programming bacterial lifestyles (planktonic, biofilm or lysis), to precisely manipulate the process of bacterial adhesion, colonization and drug release in BMCT process, via hierarchical modulation of the lighting power density (LPD) of near-infrared (NIR) light. The deep tissue penetration of NIR offers us a modality for spatiotemporal and noninvasive control of bacterial genetic circuits in vivo. By combining computational modeling with high throughput characterization device, we optimized the genetic circuits in engineered bacteria to program the process of bacterial lifestyle transitions by altering the illumination scheme of NIR. Our results showed that programming intratumoral bacterial lifestyle transitions allows precise control of multiple key steps throughout the BMCT process, and therapeutic efficacy can be greatly improved by controlling the localization and dosage of therapeutic agents via optimizing the illumination scheme.

synthetic biology↗

bioRxiv2021

Graphsite: ligand-binding site classification using deep graph neural network

Binding sites are concave surfaces on proteins that bind to small molecules called ligands. Types of molecules that bind to the protein determine its biological function. Meanwhile, the binding process between small molecules and the protein is also crucial to various biological functionalities. Therefore, identifying and classifying such binding sites would enormously contribute to biomedical applications such as drug repurposing. Deep learning is a modern artificial intelligence technology. It utilizes deep neural networks to handle complex tasks such as image classification and language translation. Previous work has proven the capability of deep learning models handle binding sites wherein the binding sites are represented as pixels or voxels. Graph neural networks (GNNs) are deep learning models that operate on graphs. GNNs are promising for handling binding sites related tasks - provided there is an adequate graph representation to model the binding sties. In this communication, we describe a GNN-based computational method, GraphSite, that utilizes a novel graph representation of ligand-binding sites. A state-of-the-art GNN model is trained to capture the intrinsic characteristics of these binding sites and classify them. Our model generalizes well to unseen data and achieves test accuracy of 81.28% on classifying 14 binding site classes.

bioinformatics↗

bioRxiv2021

3CAC: improving the classification of phages and plasmids from metagenomic assemblies using assembly graphs

MotivationBacteriophages and plasmids usually coexist with their host bacteria in microbial communities and play important roles in microbial evolution. Accurately identifying sequence contigs as phages, plasmids, and bacterial chromosomes in mixed metagenomic assemblies is critical for further unravelling their functions. Many classification tools have been developed for identifying either phages or plasmids in metagenomic assemblies. However, only two classifiers, PPR-Meta and viralVerify, were proposed to simultaneously identify phages and plasmids in mixed metagenomic assemblies. Due to the very high fraction of chromosome contigs in the assemblies, both tools achieve high precision in the classification of chromosomes but perform poorly in classifying phages and plasmids. Short contigs in these assemblies are often wrongly classified or classified as uncertain. ResultsHere we present 3CAC, a new three-class classifier that improves the precision of phage and plasmid classification. 3CAC starts with an initial three-class classification generated by existing classifiers and improves the classification of short contigs and contigs with low confidence classification by using proximity in the assembly graph. Evaluation on simulated metagenomes and on real human gut microbiome samples showed that 3CAC outperformed PPR-Meta and viralVerify in both precision and recall, and increased F1-score by 10-60 percentage points. AvailabilityThe 3CAC software is available on https://github.com/Shamir-Lab/3CAC. Contactrshamir@tau.ac.il Supplementary informationSupplementary data are available at Bioinformatics online.

bioinformatics↗

bioRxiv2021

PlasLR Enables Adaptation of Plasmid Prediction for Error-Prone Long Reads

Plasmids are extra-chromosomal genetic elements commonly found in bacterial cells that support many functional aspects including environmental adaptations. The identification of these genetic elements is vital for the further study of function and behaviour of the organisms. However it is challenging to separate these small sequences from longer chromosomes within a given species. Machine learning approaches have been successfully developed to classify assembled contigs into two classes (plasmids and chromosomes). However, such tools are not designed to directly perform classification on long and error-prone reads which have been gaining popularity in genomics studies. Assembling complete plasmids is still challenging for many long-read assemblers with a mixed input of long and error-prone reads from plasmids and chromosomes. In this paper, we present PlasLR, a tool that adapts existing plasmid detection approaches to directly classify long and error-prone reads. PlasLR makes use of both the composition and coverage information of long and error-prone reads. We evaluate PlasLR on multiple simulated and real long-read datasets with varying compositions of plasmids and chromosomes. Our experiments demonstrate that PlasLR substantially improves the accuracy of plasmid detection on top of the state-of-the-art plasmid detection tools. Moreover, we show that using PlasLR before long-read assembly helps to enhance the assembly quality in terms of plasmid recovery and near complete chromosome assembly from metagenomic datasets.

bioinformatics↗

bioRxiv2020

Longest Order Conserved Exemplar Subsequences

We propose a new problem whose input data are two linear genomes together with two indexed gene subsequences of them, which asks to find a longest common exemplar subsequence of the two given genomes with a subsequence identical to the given indexed gene subsequences. We present an algorithm for this problem such that the algorithm is allowed to take diminishing time and space to solve the problem by setting the indexed genes with an incremental number. Although an incremental number of indexed genes were selected, the algorithm was verified definite to reach a solution whose length insistently comes very close to a real longest common exemplar subsequence of the two given genomes. Aiming at 23 human/gorilla chromosome pairs, the algorithm was examined for use in questing for longest common exemplar subsequences whose basic units are annotated genes as well as pseudo genes, namely consecutive DNA subsequences. By contrasting the pseudo gene common exemplar subsequences the algorithm had reached for the human chromosomes 7 and 16 and their gorilla homologues with those annotated genes in the human and gorilla chromosomes, we found more than 1 000 and 500 pseudo genes in the human chromosomes 7 and 16 that occur in the same order as they are in the gorilla chromosomes 7 and 16 and, do not overlap with any annotated gene. Author summaryThere is a benefit of the algorithm: It can reach a long enough common exemplar subsequence of two linear genomes in as fast a speed as one requires even if the given genomes would be equipped with too many duplicated genes, which can be done by setting incremental number of indexed genes. We developed a Java software based on the algorithm, that has been available for download on https://github.com/ShuZhang-sdu/LCES. Only in need to set the indexed gene sequences as null, was it verified successful for our algorithm to obtain the longest common exemplar subsequences of the annotated gene summary pairs extracted from 23 human/gorilla chromosome pairs. In convenience for researchers to find new motifs or conserved genes, we devoted for the algorithm to quest pseudo gene (i.e. consecutive DNA subsequences) summary pairs of the 23 human/gorilla chromosome pairs for solutions. There are 20 pseudo gene summary pairs whose longest common exemplar subsequences have been found by the algorithm with null indexed gene sequences. The other 3 pseudo gene summary pairs were verified solvable for the algorithm to reach their longest common exemplar subsequences that have to admit subsequences identical to given indexed gene subsequences. There were informed to exist 2 353 and 1 148 pseudo genes in the gorilla chromosome 7 and 16 that occur in the same order as they are in the human chromosome 7 and 16 and, do not overlap with any annotated gene. These pseudo genes should be significant for annotating the human or gorilla genome.

bioinformatics↗

bioRxiv2020

Improving the efficiency of de Bruijn graph construction using compact universal hitting sets

High-throughput sequencing techniques generate large volumes of DNA sequencing data at ultra-fast speed and extremely low cost. As a consequence, sequencing techniques have become ubiquitous in biomedical research and are used in hundreds of genomic applications. Efficient data structures and algorithms have been developed to handle the large datasets produced by these techniques. The prevailing method to index DNA sequences in those data structures and algorithms is by using k-mers (k-long substrings) known as minimizers. Minimizers are the smallest k-mers selected in every consecutive window of a fixed length in a sequence, where the smallest is determined according to a predefined order, e.g., lexicographic. Recently, a new k-mer order based on a universal hitting set (UHS) was suggested. While several studies have shown that orders based on a small UHS have improved properties, the utility of using them in high-throughput sequencing analysis tasks has not been demonstrated to date. Here, we demonstrate the practical benefit of UHSs for the first time, in the genome assembly task. Reconstructing a genome from billions of short reads is a fundamental task in high-throughput sequencing analyses. De Bruijn graph construction is a key step in genome assembly, which often requires very large amounts of memory and long computation time. A critical bottleneck lies in the partitioning of DNA sequences into bins. The sequences in each bin are assembled separately, and the final de Bruijn graph is constructed by merging the bin-specific subgraphs. We incorporated a UHS-based order in the bin partition step of the Minimum Substring Partitioning algorithm of Li et al. (2013). Using a UHS-based order instead of lexicographic- or random-ordered minimizers produced lower density minimizers with more balanced bin partitioning, which led to a reduction in both runtime and memory usage.

bioinformatics↗

bioRxiv2020

Optogenetical control of infection signaling cascade of bacteria by an engineered light-responsive protein

Bacterial pathogens operate by tightly controlling the pathogenicity to facilitate invasion and survival in host. While small molecule inducers can be designed to modulate pathogenicity to perform studies of pathogen-host interaction, these approaches, due to the diffusion property of chemicals, may have unintended, or pleiotropic effects that can impose limitations on their use. By contrast, light provides superior spatial and temporal resolution. Here, using optogenetics we reengineered GacS of the opportunistic pathogen Pseudomonas aeruginosa, signal transduction protein of the global Gac/Rsm cascade which is of central importance for regulation of infection factors. The resultant protein YGS24 displayed significant light-dependent activity of GacS kinases in Pseudomonas aeruginosa. When introduced in Caenorhabditis elegans host systems, YGS24 stimulated the pathogenicity of PAO1 in BHI and of PA14 in SK medium progressively upon blue-light exposure. This optogenetic system provides an accessible way to spatiotemporally control bacterial pathogenicity in defined host even specific tissues to develop new pathogenesis systems, which may in turn expedite development of innovative therapeutics.

synthetic biology↗

bioRxiv2020

GraphGR: A graph neural network to predict the effect of pharmacotherapy on the cancer cell growth

Genomic profiles of cancer cells provide valuable information on genetic alterations in cancer. Several recent studies employed these data to predict the response of cancer cell lines to treatment with drugs. Nonetheless, due to the multifactorial phenotypes and intricate mechanisms of cancer, the accurate prediction of the effect of pharmacotherapy on a specific cell line based on the genetic information alone is problematic. High prediction accuracies reported in the literature likely result from significant overlaps among training, validation, and testing sets, making many predictors inapplicable to new data. To address these issues, we developed GraphGR, a graph neural network with sophisticated attention propagation mechanisms to predict the therapeutic effects of kinase inhibitors across various tumors. Emphasizing on the system-level complexity of cancer, GraphGR integrates multiple heterogeneous data, such as biological networks, genomics, inhibitor profiling, and genedisease associations, into a unified graph structure. In order to construct diverse and information-rich cancer-specific networks, we devised a novel graph reduction protocol based on not only the topological information, but also the biological knowledge. The performance of GraphGR, properly cross-validated at the tissue level, is 0.83 in terms of the area under the receiver operating characteristics, which is notably higher than those measured for other approaches on the same data. Finally, several new predictions are validated against the biomedical literature demonstrating that GraphGR generalizes well to unseen data, i.e. it can predict therapeutic effects across a variety of cancer cell lines and inhibitors. GraphGR is freely available to the academic community at https://github.com/pulimeng/GraphGR.

bioinformatics↗

bioRxiv2019

Radiologic Evaluation of the Influence of Cleft Type on Nasal Dorsum Growth

PurposeThe study was designed to evaluate whether intrinsic morphological characteristics of the nasal dorsum are affected by cleft type, specifically cleft lip only (CL) and cleft lip with cleft palate(CL/P).\n\nMethods576 cleft patients (278 CL only, 298 CL/P), and 333 individuals without orofacial clefts were retrospectively enrolled. Lateral cephalometric radiographs of all individuals were taken to evaluate the nasal length and nasal dorsum height. Dunns test was used to analyze the difference (p < 0.001).\n\nResultsIn CL and control, the angulation of the nasal bone and nasal dorsum increase by age similarly (5y-18y, p>0.05). In CL, the total dorsal length is significantly shorter (5y-18y, p<0.001). Although the upper nasal dorsum is similar (except in 5y-6y), the lower nasal dorsum is shorter (5y-18y, p<0.001).\n\nIn CLP, there is no significant difference in the nasal bone angle compared with controls between 5y-7y. However, it develops insufficiently as children grow (8y-18y, p<0.001). The nasal dorsum angle is notably smaller (5y-18y, p<0.001). Nasal bone length is not significantly different from control at all stages except during ages 11y-13y (p<0.05). Total nasal dorsal length is similar to the control at skeletal maturity (17y-18y, p>0.05), although it is shorter during 8y to 16y (p<0.05). The upper nasal dorsum is overdeveloped (14y-18y, p<0.05), whereas the lower nasal dorsum is underdeveloped (5y-18y, p<0.001).\n\nConclusionCL inhibits the growth of nasal dorsum length, leading to short nose deformity. CL/P patients are prone to saddle-nose deformity because of the diminished nasal height (decreased nasal angle).

developmental biology↗