scDecouple: decoupling cellular response from infected proportion bias in scCRISPR-seq.

Single-cell clustered regularly interspaced short palindromic repeats-sequencing (scCRISPR-seq) is an emerging high-throughput CRISPR screening technology where the true cellular response to perturbation is coupled with infected proportion bias of guide RNAs (gRNAs) across different cell clusters. The mixing of these effects introduces noise into scCRISPR-seq data analysis and thus obstacles to relevant studies. We developed scDecouple to decouple true cellular response of perturbation from the influence of infected proportion bias. scDecouple first models the distribution of gene expression profiles in perturbed cells and then iteratively finds the maximum likelihood of cell cluster proportions as well as the cellular response for each gRNA. We demonstrated its performance in a series of simulation experiments. By applying scDecouple to real scCRISPR-seq data, we found that scDecouple enhances the identification of biologically perturbation-related genes. scDecouple can benefit scCRISPR-seq data analysis, especially in the case of heterogeneous samples or complex gRNA libraries.

Inferring cell diversity in single cell data using consortium-scale epigenetic data as a biological anchor for cell identity.

Methods for cell clustering and gene expression from single-cell RNA sequencing (scRNA-seq) data are essential for biological interpretation of cell processes. Here, we present TRIAGE-Cluster which uses genome-wide epigenetic data from diverse bio-samples to identify genes demarcating cell diversity in scRNA-seq data. By integrating patterns of repressive chromatin deposited across diverse cell types with weighted density estimation, TRIAGE-Cluster determines cell type clusters in a 2D UMAP space. We then present TRIAGE-ParseR, a machine learning method which evaluates gene expression rank lists to define gene groups governing the identity and function of cell types. We demonstrate the utility of this two-step approach using atlases of in vivo and in vitro cell diversification and organogenesis. We also provide a web accessible dashboard for analysis and download of data and software. Collectively, genome-wide epigenetic repression provides a versatile strategy to define cell diversity and study gene regulation of scRNA-seq data.

Cellular diversity and lineage trajectory: insights from mouse single cell transcriptomes.

Single cell RNA-sequencing (scRNA-seq) technology has matured to the point that it is possible to generate large single cell atlases of developing mouse embryos. These atlases allow the dissection of developmental cell lineages and molecular changes during embryogenesis. When coupled with single cell technologies for profiling the chromatin landscape, epigenome, proteome and metabolome, and spatial tissue organisation, these scRNA-seq approaches can now collect a large volume of multi-omic data about mouse embryogenesis. In addition, advances in computational techniques have enabled the inference of developmental lineages of differentiating cells, even without explicitly introduced genetic markers. This Spotlight discusses recent advent of single cell experimental and computational methods, and key insights from applying these methods to the study of mouse embryonic development. We highlight challenges in analysing and interpreting these data to complement and expand our knowledge from traditional developmental biology studies in relation to cell identity, diversity and lineage differentiation.

Multi-Level Biomarkers for Early Diagnosis of Ischaemic Stroke: A Systematic Review and Meta-Analysis.

Blood biomarkers hold potential for the early diagnosis of ischaemic stroke (IS). We aimed to evaluate the current weight of evidence and identify potential biomarkers and biological pathways for further investigation. We searched PubMed, EMBASE, the Cochrane Library and Web of Science, used R package meta4diag for diagnostic meta-analysis and applied Gene Ontology (GO) analysis to identify vital biological processes (BPs). Among 8544 studies, we included 182 articles with a total of 30,446 participants: 15675 IS, 2317 haemorrhagic stroke (HS), 1798 stroke mimics, 846 transient ischaemic attack and 9810 control subjects. There were 518 pooled biomarkers including 203 proteins, 114 genes, 108 metabolites and 88 transcripts. Our study generated two shortlists of biomarkers for future research: one with optimal diagnostic performance and another with low selection bias. Glial fibrillary acidic protein was eligible for diagnostic meta-analysis, with summary sensitivities and specificities for differentiating HS from IS between 3 h and 24 h after stroke onset ranging from 73% to 80% and 77% to 97%, respectively. GO analysis revealed the top five BPs associated with IS. This study provides a holistic view of early diagnostic biomarkers in IS. Two shortlists of biomarkers and five BPs warrant future investigation.

Lineage tracing and single-cell analysis reveal proliferative Prom1+ tumour-propagating cells and their dynamic cellular transition during liver cancer progression.

OBJECTIVE: Hepatocellular carcinoma (HCC) has high intratumoral heterogeneity, which contributes to therapeutic resistance and tumour recurrence. We previously identified Prominin-1 (PROM1)/CD133 as an important liver cancer stem cell (CSC) marker in human HCC. The aim of this study was to investigate the heterogeneity and properties of Prom1+ cells in HCC in intact mouse models. DESIGN: We established two mouse models representing chronic fibrotic HCC and rapid steatosis-related HCC. We performed lineage tracing post-HCC induction using Prom1C-L/+; Rosa26tdTomato/+ mice, and targeted depletion using Prom1C-L/+; Rosa26DTA/+ mice. Single-cell RNA sequencing (scRNA-seq) was carried out to analyse the transcriptomic profile of traced Prom1+ cells. RESULTS: Prom1 in HCC tumours marks proliferative tumour-propagating cells with CSC-like properties. Lineage tracing demonstrated that these cells display clonal expansion in situ in primary tumours. Labelled Prom1+ cells exhibit increasing tumourigenicity in 3D culture and allotransplantation, as well as potential to form cancers of differential lineages on transplantation. Depletion of Prom1+ cells impedes tumour growth and reduces malignant cancer hallmarks in both HCC models. scRNA-seq analysis highlighted the heterogeneity of Prom1+ HCC cells, which follow a trajectory to the dedifferentiated status with high proliferation and stem cells traits. Conserved gene signature of Prom1 linage predicts poor prognosis in human HCC. The activated oxidant detoxification underlies the protective mechanism of dedifferentiated transition and lineage propagation. CONCLUSION: Our study combines in vivo lineage tracing and scRNA-seq to reveal the heterogeneity and dynamics of Prom1+ HCC cells, providing insights into the mechanistic role of malignant CSC-like cells in HCC progression.

FlowGrid enables fast clustering of very large single-cell RNA-seq data.

MOTIVATION: Scalable clustering algorithms are needed to analyze millions of cells in single cell RNA-seq (scRNA-seq) data. RESULTS: Here, we present an open source python package called FlowGrid that can integrate into the Scanpy workflow to perform clustering on very large scRNA-seq datasets. FlowGrid implements a fast density-based clustering algorithm originally designed for flow cytometry data analysis. We introduce a new automated parameter tuning procedure, and show that FlowGrid can achieve comparable clustering accuracy as state-of-the-art clustering algorithms but at a substantially reduced run time for very large single cell RNA-seq datasets. For example, FlowGrid can complete a one-hour clustering task for one million cells in about five min. AVAILABILITY AND IMPLEMENTATION: https://github.com/holab-hku/FlowGrid. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Dynamic changes in antibiotic resistance genes and gut microbiota after Helicobacter pylori eradication therapies.

BACKGROUND: Short-term antibiotics exposure is associated with alterations in microbiota and antibiotic resistance genes (ARGs) in the human gut. While antibiotics are critical in the successful eradication of Helicobacter pylori, the short-term and long-term impacts on the composition and quantity of antibiotics resistance genes after H. pylori eradication are unclear. This study used whole-genome shotgun metagenomic of stool samples to characterize the gut microbiota and ARGs, before and after H. pylori eradication therapy. RESULTS: Forty-four H. pylori-infected patients were recruited, including 21 treatment naïve patients who received clarithromycin-based triple therapy (CLA group) and 23 patients who failed previous therapies, in which 10 received levofloxacin-based quadruple therapy (LEVO group) and 13 received other combinations (OTHER group). Stool samples were collected at baseline (before current treatment), 6 week and 6 month after eradication therapy. At baseline, there was only a slight difference among the three groups on ARGs and gut microbiota. After eradication therapy, there was a transient but significant increase in gut ARGs 6 week post-therapy, among which the LEVO group had the most significant ARGs alteration compared to other two groups. For treatment naïve patients, those with higher ErmF abundance were prone to fail CLA eradication and gain more ARGs after treatment. For gut microbiota, the bacteria richness decreased at 6 week and there was a significant difference in microbiota community among the three groups at 6 week. CONCLUSIONS: Our findings demonstrated the dynamic alterations in gut microbiota and ARGs induced by different eradication therapies, which could influence the choices of antibiotics in eradication therapy.

Predicting Sepsis Mortality in a Population-Based National Database: Machine Learning Approach.

BACKGROUND: Although machine learning (ML) algorithms have been applied to point-of-care sepsis prognostication, ML has not been used to predict sepsis mortality in an administrative database. Therefore, we examined the performance of common ML algorithms in predicting sepsis mortality in adult patients with sepsis and compared it with that of the conventional context knowledge-based logistic regression approach. OBJECTIVE: The aim of this study is to examine the performance of common ML algorithms in predicting sepsis mortality in adult patients with sepsis and compare it with that of the conventional context knowledge-based logistic regression approach. METHODS: We examined inpatient admissions for sepsis in the US National Inpatient Sample using hospitalizations in 2010-2013 as the training data set. We developed four ML models to predict in-hospital mortality: logistic regression with least absolute shrinkage and selection operator regularization, random forest, gradient-boosted decision tree, and deep neural network. To estimate their performance, we compared our models with the Super Learner model. Using hospitalizations in 2014 as the testing data set, we examined the models' area under the receiver operating characteristic curve (AUC), confusion matrix results, and net reclassification improvement. RESULTS: Hospitalizations of 923,759 adults were included in the analysis. Compared with the reference logistic regression (AUC: 0.786, 95% CI 0.783-0.788), all ML models showed superior discriminative ability (P<.001), including logistic regression with least absolute shrinkage and selection operator regularization (AUC: 0.878, 95% CI 0.876-0.879), random forest (AUC: 0.878, 95% CI 0.877-0.880), xgboost (AUC: 0.888, 95% CI 0.886-0.889), and neural network (AUC: 0.893, 95% CI 0.891-0.895). All 4 ML models showed higher sensitivity, specificity, positive predictive value, and negative predictive value compared with the reference logistic regression model (P<.001). We obtained similar results from the Super Learner model (AUC: 0.883, 95% CI 0.881-0.885). CONCLUSIONS: ML approaches can improve sensitivity, specificity, positive predictive value, negative predictive value, discrimination, and calibration in predicting in-hospital mortality in patients hospitalized with sepsis in the United States. These models need further validation and could be applied to develop more accurate models to compare risk-standardized mortality rates across hospitals and geographic regions, paving the way for research and policy initiatives studying disparities in sepsis care.

Autologous bone fragments for skull reconstruction after microvascular decompression.

BACKGROUND: Various methods are used to reconstruct the skull after microvascular decompression, giving their own advantages and disadvantages. The objective of this study was to evaluate the efficacy of using autologous bone fragments for skull reconstruction after microvascular decompression. METHODS: The clinical and follow-up data of 145 patients who underwent microvascular decompression and skull reconstruction using autologous bone fragments in our hospital from September 2020 to September 2021 were retrospectively analyzed. RESULTS: Three patients (2.06%) had delayed wound healing after surgery and were discharged after wound cleaning. No patient developed postoperative cerebrospinal fluid leakage, incisional dehiscence, or intracranial infection. Eighty-five (58.62%) patients underwent follow-up cranial computed tomography at 1 year postoperatively, showed excellent skull reconstruction. And, the longer the follow-up period, the more satisfactory the cranial repair. Two patients underwent re-operation for recurrence of hemifacial spasm, and intraoperative observation revealed that the initial skull defect was filled with new skull bone. CONCLUSION: The use of autologous bone fragments for skull reconstruction after microvascular decompression is safe and feasible, with few postoperative wound complications and excellent long-term repair results.

IgA binds to the AD-2 epitope of glycoprotein B and neutralizes human cytomegalovirus.

Human cytomegalovirus (HCMV) is a ubiquitous pathogen that is potentially pathogenic in immunosuppressed individuals and pregnant females during primary infection. The HCMV envelope glycoprotein B (gB) facilitates viral entry into all cell types and induces a potent immune response. AD-2 epitope is a highly conserved linear neutralizing epitope of gB and a critical target for antibodies; however, only 50% of sero-positive individuals make IgG antibodies to this site and IgA responses have not been fully investigated. This study aimed to compare IgG and IgA responses against gB and the AD-2 epitope in naturally exposed individuals and those receiving a recombinant gB/MF59 adjuvant vaccine. Thus, vaccination of sero-positive individuals improved pre-existing gB-specific IgA and IgG levels and induced de novo gB-specific IgA and IgG responses in sero-negative recipients. Pre-existing AD-2 IgG and IgA responses were boosted with vaccination, but de novo AD-2 responses were not detected. Naturally exposed individuals had dominant IgG responses towards gB and AD-2 compared with weaker and variable IgA responses, although a significant IgA binding response to AD-2 was observed within human breastmilk samples. All antibodies binding AD-2 contained kappa light chains, whereas balanced kappa/lambda light chain usage was found for those binding to gB. V region-matched AD-2-specific recombinant IgG and IgA bound both to gB and to AD-2 and neutralized HCMV infection in vitro. Overall, these results indicate that although human IgG responses dominate, IgA class antibodies against AD-2 are a significant component of human milk, which may function to protect neonates from HCMV.

Non-self mutation: double-stranded RNA elicits antiviral pathogenic response in a Drosophila model of expanded CAG repeat neurodegenerative diseases.

Inflammation is activated prior to symptoms in neurodegenerative diseases, providing a plausible pathogenic mechanism. Indeed, genetic and pharmacological ablation studies in animal models of several neurodegenerative diseases demonstrate that inflammation is required for pathology. However, while there is growing evidence that inflammation-mediated pathology may be the common mechanism underlying neurodegenerative diseases, including those due to dominantly inherited expanded repeats, the proximal causal agent is unknown. Expanded CAG.CUG repeat double-stranded RNA causes inflammation-mediated pathology when expressed in Drosophila. Repeat dsRNA is recognized by Dicer-2 as a foreign or 'non-self' molecule triggering both antiviral RNA and RNAi pathways. Neither of the RNAi pathway cofactors R2D2 nor loquacious are necessary, indicating antiviral RNA activation. RNA modification enables avoidance of recognition as 'non-self' by the innate inflammatory surveillance system. Human ADAR1 edits RNA conferring 'self' status and when co-expressed with expanded CAG.CUG dsRNA in Drosophila the pathology is lost. Cricket Paralysis Virus protein CrPV-1A is a known antagonist of Argonaute-2 in Drosophila antiviral defense. CrPV-1A co-expression also rescues pathogenesis, confirming anti-viral-RNA response. Repeat expansion mutation therefore confers 'non-self' recognition of endogenous RNA, thereby providing a proximal, autoinflammatory trigger for expanded repeat neurodegenerative diseases.

Light-focusing human micro-lenses generated from pluripotent stem cells model lens development and drug-induced cataract in vitro.

Cataracts cause vision loss and blindness by impairing the ability of the ocular lens to focus light onto the retina. Various cataract risk factors have been identified, including drug treatments, age, smoking and diabetes. However, the molecular events responsible for these different forms of cataract are ill-defined, and the advent of modern cataract surgery in the 1960s virtually eliminated access to human lenses for research. Here, we demonstrate large-scale production of light-focusing human micro-lenses from spheroidal masses of human lens epithelial cells purified from differentiating pluripotent stem cells. The purified lens cells and micro-lenses display similar morphology, cellular arrangement, mRNA expression and protein expression to human lens cells and lenses. Exposing the micro-lenses to the emergent cystic fibrosis drug Vx-770 reduces micro-lens transparency and focusing ability. These human micro-lenses provide a powerful and large-scale platform for defining molecular disease mechanisms caused by cataract risk factors, for anti-cataract drug screening and for clinically relevant toxicity assays.

dv-trio: a family-based variant calling pipeline using DeepVariant.

MOTIVATION: In 2018, Google published an innovative variant caller, DeepVariant, which converts pileups of sequence reads into images and uses a deep neural network to identify single-nucleotide variants and small insertion/deletions from next-generation sequencing data. This approach outperforms existing state-of-the-art tools. However, DeepVariant was designed to call variants within a single sample. In disease sequencing studies, the ability to examine a family trio (father-mother-affected child) provides greater power for disease mutation discovery. RESULTS: To further improve DeepVariant's variant calling accuracy in family-based sequencing studies, we have developed a family-based variant calling pipeline, dv-trio, which incorporates the trio information from the Mendelian genetic model into variant calling based on DeepVariant. AVAILABILITY AND IMPLEMENTATION: dv-trio is available via an open source BSD3 license at GitHub (https://github.com/VCCRI/dv-trio/). CONTACT: e.giannoulatou@victorchang.edu.au. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PARC: ultrafast and accurate clustering of phenotypic data of millions of single cells.

MOTIVATION: New single-cell technologies continue to fuel the explosive growth in the scale of heterogeneous single-cell data. However, existing computational methods are inadequately scalable to large datasets and therefore cannot uncover the complex cellular heterogeneity. RESULTS: We introduce a highly scalable graph-based clustering algorithm PARC-Phenotyping by Accelerated Refined Community-partitioning-for large-scale, high-dimensional single-cell data (>1 million cells). Using large single-cell flow and mass cytometry, RNA-seq and imaging-based biophysical data, we demonstrate that PARC consistently outperforms state-of-the-art clustering algorithms without subsampling of cells, including Phenograph, FlowSOM and Flock, in terms of both speed and ability to robustly detect rare cell populations. For example, PARC can cluster a single-cell dataset of 1.1 million cells within 13 min, compared with >2 h for the next fastest graph-clustering algorithm. Our work presents a scalable algorithm to cope with increasingly large-scale single-cell analysis. AVAILABILITY AND IMPLEMENTATION: https://github.com/ShobiStassen/PARC. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The short isoform of the CEACAM1 receptor in intestinal T cells regulates mucosal immunity and homeostasis via Tfh cell induction.

Carcinoembryonic antigen cell adhesion molecule like I (CEACAM1) is expressed on activated T cells and signals through either a long (L) cytoplasmic tail containing immune receptor tyrosine based inhibitory motifs, which provide inhibitory function, or a short (S) cytoplasmic tail with an unknown role. Previous studies on peripheral T cells show that CEACAM1-L isoforms predominate with little to no detectable CEACAM1-S isoforms in mouse and human. We show here that this was not the case in tissue resident T cells of intestines and gut associated lymphoid tissues, which demonstrated predominant expression of CEACAM1-S isoforms relative to CEACAM1-L isoforms in human and mouse. This tissue resident predominance of CEACAM1-S expression was determined by the intestinal environment where it served a stimulatory function leading to the regulation of T cell subsets associated with the generation of secretory IgA immunity, the regulation of mucosal commensalism, and defense of the barrier against enteropathogens.

Ularcirc: visualization and enhanced analysis of circular RNAs via back and canonical forward splicing.

Circular RNAs (circRNA) are a unique class of transcripts that can only be identified from sequence alignments spanning discordant junctions, commonly referred to as backsplice junctions (BSJ). Canonical splicing is also linked with circRNA biogenesis either from the parental transcript or internal to the circRNA, and is not fully utilized in circRNA software. Here we present Ularcirc, a software tool that integrates the visualization of both BSJ and forward splicing junctions and provides downstream analysis of selected circRNA candidates. Ularcirc utilizes the output of CIRI, circExplorer, or raw chimeric output of the STAR aligner and assembles BSJ count table to allow multi-sample analysis. We used Ularcirc to identify and characterize circRNA from public and in-house generated data sets and demonstrate how it can be used to (i) discover novel splicing patterns of parental transcripts, (ii) detect internal splicing patterns of circRNA, and (iii) reveal the complexity of BSJ formation. Furthermore, we identify circRNA that have potential open reading frames longer than their linear sequence. Finally, we detected and validated the presence of a novel class of circRNA generated from ApoA4 transcripts whose BSJ derive from multiple non-canonical splicing sites within coding exons. Ularcirc is accessed via https://github.com/VCCRI/Ularcirc.

NAD Deficiency, Congenital Malformations, and Niacin Supplementation.

BACKGROUND: Congenital malformations can be manifested as combinations of phenotypes that co-occur more often than expected by chance. In many such cases, it has proved difficult to identify a genetic cause. We sought the genetic cause of cardiac, vertebral, and renal defects, among others, in unrelated patients. METHODS: We used genomic sequencing to identify potentially pathogenic gene variants in families in which a person had multiple congenital malformations. We tested the function of the variant by using assays of in vitro enzyme activity and by quantifying metabolites in patient plasma. We engineered mouse models with similar variants using the CRISPR (clustered regularly interspaced short palindromic repeats)-Cas9 system. RESULTS: Variants were identified in two genes that encode enzymes of the kynurenine pathway, 3-hydroxyanthranilic acid 3,4-dioxygenase (HAAO) and kynureninase (KYNU). Three patients carried homozygous variants predicting loss-of-function changes in the HAAO or KYNU proteins (HAAO p.D162*, HAAO p.W186*, or KYNU p.V57Efs*21). Another patient carried heterozygous KYNU variants (p.Y156* and p.F349Kfs*4). The mutant enzymes had greatly reduced activity in vitro. Nicotinamide adenine dinucleotide (NAD) is synthesized de novo from tryptophan through the kynurenine pathway. The patients had reduced levels of circulating NAD. Defects similar to those in the patients developed in the embryos of Haao-null or Kynu-null mice owing to NAD deficiency. In null mice, the prevention of NAD deficiency during gestation averted defects. CONCLUSIONS: Disruption of NAD synthesis caused a deficiency of NAD and congenital malformations in humans and mice. Niacin supplementation during gestation prevented the malformations in mice. (Funded by the National Health and Medical Research Council of Australia and others.).

Multi-omic profiling reveals associations between the gut mucosal microbiome, the metabolome, and host DNA methylation associated gene expression in patients with colorectal cancer.

BACKGROUND: The human gut microbiome plays a critical role in the carcinogenesis of colorectal cancer (CRC). However, a comprehensive analysis of the interaction between the host and microbiome is still lacking. RESULTS: We found correlations between the change in abundance of microbial taxa, butyrate-related colonic metabolites, and methylation-associated host gene expression in colonic tumour mucosa tissues compared with the adjacent normal mucosa tissues. The increase of genus Fusobacterium abundance was correlated with a decrease in the level of 4-hydroxybutyric acid (4-HB) and expression of immune-related peptidase inhibitor 16 (PI16), Fc Receptor Like A (FCRLA) and Lymphocyte Specific Protein 1 (LSP1). The decrease in the abundance of another potentially 4-HB-associated genus, Prevotella 2, was also found to be correlated with the down-regulated expression of metallothionein 1 M (MT1M). Additionally, the increase of glutamic acid-related family Halomonadaceae was correlated with the decreased expression of reelin (RELN). The decreased abundance of genus Paeniclostridium and genus Enterococcus were correlated with increased lactic acid level, and were also linked to the expression change of Phospholipase C Beta 1 (PLCB1) and Immunoglobulin Superfamily Member 9 (IGSF9) respectively. Interestingly, 4-HB, glutamic acid and lactic acid are all butyrate precursors, which may modify gene expression by epigenetic regulation such as DNA methylation. CONCLUSIONS: Our study identified associations between previously reported CRC-related microbial taxa, butyrate-related metabolites and DNA methylation-associated gene expression in tumour and normal colonic mucosa tissues from CRC patients, which uncovered a possible mechanism of the role of microbiome in the carcinogenesis of CRC. In addition, these findings offer insight into potential new biomarkers, therapeutic and/or prevention strategies for CRC.

Effect of machine learning re-sampling techniques for imbalanced datasets in 18F-FDG PET-based radiomics model on prognostication performance in cohorts of head and neck cancer patients.

PURPOSE: Biomedical data frequently contain imbalance characteristics which make achieving good predictive performance with data-driven machine learning approaches a challenging task. In this study, we investigated the impact of re-sampling techniques for imbalanced datasets in PET radiomics-based prognostication model in head and neck (HNC) cancer patients. METHODS: Radiomics analysis was performed in two cohorts of patients, including 166 patients newly diagnosed with nasopharyngeal carcinoma (NPC) in our centre and 182 HNC patients from open database. Conventional PET parameters and robust radiomics features were extracted for correlation analysis of the overall survival (OS) and disease progression-free survival (DFS). We investigated a cross-combination of 10 re-sampling methods (oversampling, undersampling, and hybrid sampling) with 4 machine learning classifiers for survival prediction. Diagnostic performance was assessed in hold-out test sets. Statistical differences were analysed using Monte Carlo cross-validations by post hoc Nemenyi analysis. RESULTS: Oversampling techniques like ADASYN and SMOTE could improve prediction performance in terms of G-mean and F-measures in minority class, without significant loss of F-measures in majority class. We identified optimal PET radiomics-based prediction model of OS (AUC of 0.82, G-mean of 0.77) for our NPC cohort. Similar findings that oversampling techniques improved the prediction performance were seen when this was tested on an external dataset indicating generalisability. CONCLUSION: Our study showed a significant positive impact on the prediction performance in imbalanced datasets by applying re-sampling techniques. We have created an open-source solution for automated calculations and comparisons of multiple re-sampling techniques and machine learning classifiers for easy replication in future studies.

Mass spectrometry studies of the fragmentation patterns and mechanisms of protonated peptoids.

Peptoids belong to a class of sequence-controlled polymers comprising of N-alkylglycine. This study focuses on using tandem mass spectrometry techniques to characterize the fragmentation patterns of a set of singly and doubly protonated peptoids consisting of one basic residue placed at different positions. The singly protonated peptoids fragment by producing predominately high-abundant C-terminal ions called Y-ions and low-abundant N-terminal ions called B-ions. Computational studies suggest that the proton affinity (PA) of the C-terminal fragments is generally higher than that of the N-terminal fragments, and the PA of the former increases as the fragments are elongated. The B-ions are likely formed upon dissociating the proton-activated amide bonds via an oxazolone structure, and the Y-ions are produced subsequently by abstracting a proton from the newly formed B-ions, which is energetically favored. The doubly protonated peptoids prefer to fragment closest to either the N- or the C-terminus and produce corresponding B/Y-ion pairs. The basic residue seems to dictate the preferred fragmentation site, which may be the result of minimizing the repulsion between the two charges. Water and terminal neutral losses are a facile process accompanying the peptoid fragmentation in both charge states. The patterns appear to be highly influenced by the location of the basic residue.