CSChayaporn Suphavilai

Chayaporn Suphavilai

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Chayaporn Suphavilai has a total of 156 co-authors. Here are the top co-authors by number of shared publications:

  • May Kyawt Aung

    Department of Infection Prevention and Epidemiology, Singapore General Hospital, Singapore

    4 shared publications
  • Niranjan Nagarajan

    Genome Institute of Singapore, Agency for Science, Technology and Research (A*STAR), Singapore, Singapore

    4 shared publications
  • Indumathi Venkatachalam

    Department of Infection Prevention and Epidemiology, Singapore General Hospital, Singapore; Department of Infectious Diseases, Singapore General Hospital, Singapore; Duke-NUS Medical School, Singapore

    3 shared publications
  • Karrie Kwan Ki Ko

    Department of Microbiology, Singapore General Hospital, Singapore, Singapore

    3 shared publications
  • Kwee Yuen Tan

    Department of Infection Prevention and Epidemiology, Singapore General Hospital, Singapore

    3 shared publications
  • Lai Chee Lee

    Department of Infection Prevention and Epidemiology, Singapore General Hospital, Singapore

    3 shared publications

Latest Publications

  • Top Cited Papers
    Open Access

    Detection and characterisation of a sixth Candida auris clade in Singapore: a genomic and phenotypic study

    by Chayaporn Suphavilai, Chayaporn Suphavilai, Chayaporn Suphavilai, Karrie Kwan Ki Ko, Karrie Kwan Ki Ko, Karrie Kwan Ki Ko, Kar Mun Lim, Kar Mun Lim, Kar Mun Lim, Mei Gie Tan, et al.


  • Open Access

    Addressing pandemic-wide systematic errors in the SARS-CoV-2 phylogeny

    by Martin Hunt, Angie S. Hinrichs, Daniel Anderson, Lily Karim, Bethany L Dearlove, Jeff Knaggs, Bede Constantinides, Philip W. Fowler, Gillian Rodger, Teresa Street, et al.

    The SARS-CoV-2 genome occupies a unique place in infection biology – it is the most highly sequenced genome on earth (making up over 20% of public sequencing datasets) with fine scale information on sampling date and geography, and has been subject to unprecedented intense analysis. As a result, these phylogenetic data are an incredibly valuable resource for science and public health. However, the vast majority of the data was sequenced by tiling amplicons across the full genome, with amplicon schemes that changed over the pandemic as mutations in the viral genome interacted with primer binding sites. In combination with the disparate set of genome assembly workflows and lack of consistent quality control (QC) processes, the current genomes have many systematic errors that have evolved with the virus and amplicon schemes. These errors have significant impacts on the phylogeny, and therefore over the last few years, many thousands of hours of researchers time has been spent in “eyeballing” trees, looking for artefacts, and then patching the tree. Given the huge value of this dataset, we therefore set out to reprocess the complete set of public raw sequence data in a rigorous amplicon-aware manner, and build a cleaner phylogeny. Here we provide a global tree of 4,471,579 samples, built from a consistently assembled set of high quality consensus sequences from all available public data as of June 2024, viewable at https://viridian.taxonium.org . Each genome was constructed using a novel assembly tool called Viridian ( https://github.com/iqbal-lab-org/viridian ), developed specifically to process amplicon sequence data, eliminating artefactual errors and mask the genome at low quality positions. We provide simulation and empirical validation of the methodology, and quantify the improvement in the phylogeny. We hope the tree, consensus sequences and Viridian will be a valuable resource for researchers.


  • Open Access

    MetageNN: a memory-efficient neural network taxonomic classifier robust to sequencing errors and missing genomes

    by Rafael Peres da Silva, Chayaporn Suphavilai, Niranjan Nagarajan

    Background: With the rapid increase in throughput of long-read sequencing technologies, recent studies have explored their potential for taxonomic classification by using alignment-based approaches to reduce the impact of higher sequencing error rates. While alignment-based methods are generally slower, k-mer-based taxonomic classifiers can overcome this limitation, potentially at the expense of lower sensitivity for strains and species that are not in the database. Results: We present MetageNN, a memory-efficient long-read taxonomic classifier that is robust to sequencing errors and missing genomes. MetageNN is a neural network model that uses short k-mer profiles of sequences to reduce the impact of distribution shifts on error-prone long reads. Benchmarking MetageNN against other machine learning approaches for taxonomic classification (GeNet) showed substantial improvements with long-read data (20% improvement in F1 score). By utilizing nanopore sequencing data, MetageNN exhibits improved sensitivity in situations where the reference database is incomplete. It surpasses the alignment-based MetaMaps and MEGAN-LR, as well as the k-mer-based Kraken2 tools, with improvements of 100%, 36%, and 23% respectively at the read-level analysis. Notably, at the community level, MetageNN consistently demonstrated higher sensitivities than the previously mentioned tools. Furthermore, MetageNN requires < 1 / 4 th of the database storage used by Kraken2, MEGAN-LR and MMseqs2 and is >7x faster than MetaMaps and GeNet and >2x faster than MEGAN-LR and MMseqs2. Conclusion: This proof of concept work demonstrates the utility of machine-learning-based methods for taxonomic classification using long reads. MetageNN can be used on sequences not classified by conventional methods and offers an alternative approach for memory-efficient classifiers that can be optimized further.

  • Open Access

    Discovery of the sixth Candida auris clade in Singapore

    by Chayaporn Suphavilai, Karrie Kwan Ki Ko, Kar Mun Lim, Mei Gie Tan, Patipan Boonsimma, Joash Jun Keat Chu, Sui Sin Goh, Prevena Rajandran, Lai Chee Lee, Kwee Yuen Tan, et al.

    Summary: Background: The emerging fungal pathogen Candida auris poses a serious threat to global public health due to its worldwide distribution, multidrug-resistance, high transmissibility, propensity to cause outbreaks and high mortality rates. We report three C. auris isolates detected in Singapore, which are genetically distinct from all known clades (I-V) and represent a new clade (Clade VI). Methods: Three epidemiologically unlinked clinical isolates belonging to the potential new C. auris clade were whole-genome sequenced and phenotypically characterized. The complete genomes of these isolates were compared to representative genomes of all known clades. To provide a global context, 3,651 international whole-genome sequences (WGS) from the NCBI database were included in the high-resolution single nucleotide polymorphism (SNP) analysis. Antifungal resistance genes, mating type locus, and chromosomal rearrangements were characterized from the WGS data of the Clade VI isolates. We further implemented Bayesian logistic regression models to simulate the automatic detection of Clade V and VI as their WGS data became available. Findings: The three Clade VI isolates were separated by >36,000 SNPs from all existing C. auris clades. These isolates had opposite mating type allele and different chromosomal rearrangements when compared to their closest Clade IV relatives. As a proof-of-concept, our classification model was able to flag these outlier genomes as a potential new clade. Furthermore, an independent WGS submission from Bangladesh was found to belong to this new clade. Interpretation: The discovery of a new C. auris clade in Singapore and Bangladesh, showing close relationship to Clade IV members in South America, highlights the unknown genetic diversity and origin of C. auris , particularly in under-resourced regions. Active surveillance in clinical settings, along with effective sequencing strategies and downstream analysis, will be essential in the identification of novel strains, tracking of transmission, and containment of adverse clinical impacts caused by C. auris infections. Funding: This work was supported by the Singapore National Medical Research Council (NMRC) research training fellowship (MOH-FLWSHP19may-0005), the NCRS Duke-NUS Academic Medical Center Academic Clinical Program grant (09/FY2022/P1/17-A32, GRDUKP003401), and the Genedant-GIS Innovation Program grant. Research in context: Evidence before this study: We searched PubMed using the search terms “ Candida auris ” AND “clade”, for papers published between Jan 1, 2009, and July 1, 2023. This search retrieved 115 publications. 60 relevant publications were identified. 28 studies analyzed and discussed the molecular epidemiology of C. auris , including the description of C. auris clades, either in outbreak or surveillance settings. There were 11 case reports of C. auris clinical cases that included clade determination. Two studies focused on the detection and clade determination of C. auris from non-healthcare environments. Clade-specific characteristics were described or analyzed in 14 studies. One study applied machine learning to C. auris drug resistance analysis, but not for clade determination. Four studies focused on the description of potentially new C. auris lineages, subclades, or clades. All publications described isolates that belong to one of the five known C. auris clades (I-V). All publications found that strains from different clades differed by more than 35,000 SNPs, and that there are clade-specific differences in geographical distribution, phenotypic characteristics, antifungal susceptibility profile, outbreak potential, and clinical manifestations. The NCBI Pathogen Detection system contained 4,506 C. auris genomes on July 1, 2023. There were ten (0·22%) submissions from Southeast Asian countries and 92 (2·04%) submissions from South Asia and the Indian subcontinent, which are parts of the Indomalayan biogeographic realm. Added value of this study: To the best of our knowledge, we are the first group to perform hybrid assemblies on three representative isolates in a new C. auris clade, which is separated from all other existing clades (I-V) by >36,000 SNPs. Whole-genome SNP analysis and phenotypic characterization of these epidemiologically unlinked isolates detected in Singapore suggest that they represent a previously unreported sixth major clade. High-resolution SNP analysis of 3,651 international whole-genome sequences from the NCBI database, which generated the final dataset consisting of more than 6.6 million genome pairs, revealed six distinct genetic clusters representing the five known clades and the new sixth clade (Indomalayan). In addition, we demonstrate that a machine learning approach can be used to flag these outlier genomes for further investigations as soon as they become available, thus providing the earliest possible alert for potential new public health threats. Implications of all the available evidence: Despite the high antimicrobial resistance burden in Southeast Asia and South Asia, these regions are disproportionately underrepresented in terms of genomic surveillance of C. auris , a multidrug-resistant fungal pathogen. The detection of three epidemiologically unlinked C. auris isolates in Singapore belonging to a new C. auris clade suggests that yet-to-be-reported strains may be circulating in the region. Given the propensity for multidrug resistance, healthcare-associated infection outbreaks, and the associated high mortality, active surveillance and continued vigilance is necessary.


  • Open Access

    CREAMMIST: an integrative probabilistic database for cancer drug response prediction

    by Hatairat Yingtaweesittikul, Jiaxi Wu, Aanchal Mongia, Rafael Peres, Karrie Ko, Niranjan Nagarajan, Chayaporn Suphavilai

    Extensive in vitro cancer drug screening datasets have enabled scientists to identify biomarkers and develop machine learning models for predicting drug sensitivity. While most advancements have focused on omics profiles, cancer drug sensitivity scores precalculated by the original sources are often used as-is, without consideration for variabilities between studies. It is well-known that significant inconsistencies exist between the drug sensitivity scores across datasets due to differences in experimental setups and preprocessing methods used to obtain the sensitivity scores. As a result, many studies opt to focus only on a single dataset, leading to underutilization of available data and a limited interpretation of cancer pharmacogenomics analysis. To overcome these caveats, we have developed CREAMMIST (https://creammist.mtms.dev), an integrative database that enables users to obtain an integrative dose-response curve, to capture uncertainty (or high certainty when multiple datasets well align) across five widely used cancer cell-line drug–response datasets. We utilized the Bayesian framework to systematically integrate all available dose-response values across datasets (>14 millions dose-response data points). CREAMMIST provides easy-to-use statistics derived from the integrative dose-response curves for various downstream analyses such as identifying biomarkers, selecting drug concentrations for experiments, and training robust machine learning models.


  • Nosocomial severe acute respiratory coronavirus virus 2 (SARS-CoV-2) transmission arising from a case of N-gene dropout on reverse-transcription polymerase chain reaction (RT-PCR) testing

    by Liang En Ian Wee, Karrie Kwan-Ki Ko, Edwin Philip Conceicao, May Kyawt Aung, Myat Oo Aung, Yong Yang, Shalvi Arora, Chayaporn Suphavilai, Indumathi Venkatachalam


  • Open Access

    Linking sporadic hospital clusters during a community surge of the severe acute respiratory coronavirus virus 2 (SARS-CoV-2) B.1.617.2 delta variant: The utility of whole-genome sequencing

    by Liang En Wee, Karrie Kwan-Ki Ko, Edwin Philip Conceicao, Jean Xiang-Ying Sim, Nurdyana Abdul Rahman, Shireen Yan-Ling Tan, Delphine Yan-Hong Cao, Kenneth Xing-Long Chan, May Kyawt Aung, Myat Oo Aung, et al.

    Sporadic clusters of healthcare-associated coronavirus disease 2019 (COVID-19) occurred despite intense rostered routine surveillance and a highly vaccinated healthcare worker (HCW) population, during a community surge of the severe acute respiratory coronavirus virus 2 (SARS-CoV-2) B.1.617.2 δ (delta) variant. Genomic analysis facilitated timely cluster detection and uncovered additional linkages via HCWs moving between clinical areas and among HCWs sharing a common lunch area, enabling early intervention.


  • Open Access

    Network-guided supervised learning on gene expression using a graph convolutional neural network

    by Hatairat Yingtaweesittikul, Chayaporn Suphavilai

    Background: Transcriptomic profiles have become crucial information in understanding diseases and improving treatments. While dysregulated gene sets are identified via pathway analysis, various machine learning models have been proposed for predicting phenotypes such as disease type and drug response based on gene expression patterns. However, these models still lack interpretability, as well as the ability to integrate prior knowledge from a protein-protein interaction network. Results: We propose Grandline , a graph convolutional neural network that can integrate gene expression data and structure of the protein interaction network to predict a specific phenotype. Transforming the interaction network into a spectral domain enables convolution of neighbouring genes and pinpointing high-impact subnetworks, which allow better interpretability of deep learning models. Grandline achieves high phenotype prediction accuracy (67-85% in 8 use cases), comparable to state-of-the-art machine learning models while requiring a smaller number of parameters, allowing it to learn complex but interpretable gene expression patterns from biological datasets. Conclusion: To improve the interpretability of phenotype prediction based on gene expression patterns, we developed Grandline using graph convolutional neural network technique to integrate protein interaction information. We focus on improving the ability to learn nonlinear relationships between gene expression patterns and a given phenotype and incorporation of prior knowledge, which are the main challenges of machine learning models for biological datasets. The graph convolution allows us to aggregate information from relevant genes and reduces the number of trainable parameters, facilitating model training for a small-sized biological dataset.


  • Open Access

    CalmBelt: Rapid SARS-CoV-2 Genome Characterization for Outbreak Tracking

    by Hatairat Yingtaweesittikul, Karrie Ko, Nurdyana Abdul Rahman, Shireen Yan Ling Tan, Niranjan Nagarajan, Chayaporn Suphavilai

    Background: The ongoing COVID-19 pandemic is a global health crisis caused by the spread of SARS-CoV-2. Establishing links between known cases is crucial for the containment of COVID-19. In the healthcare setting, the ability to rapidly identify potential healthcare-associated COVID-19 clusters is critical for healthcare worker and patient safety. Increasing sequencing technology accessibility has allowed routine clinical diagnostic laboratories to sequence SARS-CoV-2 in clinical samples. However, these laboratories often lack specialized informatics skills required for sequence analysis. Therefore, an on-site, intuitive sequence analysis tool that enables clinical laboratory users to analyze multiple genomes and derive clinically relevant information within an actionable timeframe is needed.Results: We propose CalmBelt, an integrated framework for on-site whole genome characterization and outbreak tracking. Nanopore sequencing technology enables on-site sequencing and construction of draft genomes for multiple SARS-CoV-2 samples within 12 h. CalmBelt's interactive interface allows users to analyse multiple SARS-CoV-2 genomes by utilizing whole genome information, collection date, and additional information such as predefined potential clusters from epidemiological investigations. CalmBelt also integrates established SARS-CoV-2 nomenclature assignments, GISAID clades and PANGO lineages, allowing users to visualize relatedness between samples together with the nomenclatures. We demonstrated multiple use cases including investigation of potential hospital transmission, mining transmission patterns in a large outbreak, and monitoring possible diagnostic-escape.Conclusions: This paper presents an on-site rapid framework for SARS-CoV-2 whole genome characterization. CalmBelt interactive web application allows non-technical users, such as routine clinical laboratory users in hospitals to determine SARS-CoV-2 variants of concern, as well as investigate the presence of potential transmission clusters. The framework is designed to be compatible with routine usage in clinical laboratories as it only requires readily available sample data, and generates information that impacts immediate infection control mitigations.


  • Open Access

    TUGDA: Task uncertainty guided domain adaptation for robust generalization of cancer drug response prediction from in vitro to in vivo settings

    by Rafael Peres da Silva, Chayaporn Suphavilai, Niranjan Nagarajan

    Over the last decade, large-scale cancer omics studies have highlighted the diversity of patient molecular profiles and the importance of leveraging this information to deliver the right drug to the right patient at the right time. Key challenges in learning predictive models for this include the high-dimensionality of omics data, limitations in the number of data points available, and heterogeneity in biological and clinical factors affecting patient response. The use of multi-task learning (MTL) techniques has been widely explored to address dataset limitations for in vitro drug response models, while domain adaptation (DA) has been employed to extend them to predict in vivo response. In both of these transfer learning settings, noisy data for some tasks (or domains) can substantially reduce the performance for others compared to single-task (domain) learners, i.e. lead to negative transfer (NT). We describe a novel multi-task unsupervised domain adaptation method (TUGDA) that addresses these limitations in a unified framework by quantifying uncertainty in predictors and weighting their influence on shared domain/task feature representations. TUGDA’s ability to rely more on predictors with low-uncertainty allowed it to notably reduce cases of negative transfer for in vitro models (63% for drugs with limited data and 94% overall) compared to state-of-the-art methods. For domain adaptation to in vivo settings, TUGDA improved performance for 6 out of 12 drugs in patient-derived xenografts, and 7 out of 22 drugs in TCGA patient datasets, despite being trained in an unsupervised fashion. TUGDA’s ability to avoid negative transfer thus provides a key capability as we try to integrate diverse drug-response datasets to build consistent predictive models with in vivo utility. Availability: https://github.com/CSB5/TUGDA