Biology subjects

Tran, D. H.

Publications and source records attributed to Tran, D. H..

8 recordsLinked to original sources

bioRxiv2026

Development and cross-cohort evaluation of a multimodal deep-learning model for ER, PR and HER2 status classification in breast cancer

Background. Estrogen receptor (ER), progesterone receptor (PR) and HER2 status guide systemic treatment in breast cancer. Multimodal models are rarely benchmarked against key biomarker molecular baselines or evaluated for cross-cohort calibration. We assessed incremental value, dependence on label-proximal HER2 transcripts, and transport of discrimination and calibration. Methods. A model combining whole-slide images, transcriptomes, age and stage was developed in TCGA-BRCA (1051 patients; ER/PR/HER2 analysis sets: 1001/998/1006) and evaluated in CPTAC-BRCA (120 patients; 118/113/95). CPTAC had been observed during earlier development and included 13 HER2-positive cases. Analyses included paired model comparisons, transcript removal, alternative loss functions and recalibration. Internal discrimination used pooled out-of-fold predictions; external discrimination used fold-mean estimates, with ensemble calibration reported separately. Internal calibration and macro-AUC uncertainty were interpreted as exploratory. Results. Internal AUCs were 0.943 for ER, 0.901 for PR, and 0.860 for HER2; external AUCs were 0.952, 0.937, and 0.918, respectively. External macro-AUC was 0.936 for the multimodal model and 0.957 for logistic regression using ESR1, PGR, and ERBB2 (difference -0.021). Removal of the extended 29-transcript 17q12-q21 region changed HER2 AUC by -0.035 [-0.058 to -0.013] (95% CI) internally and by -0.042 [-0.089 to -0.007] externally relative to the baseline multimodal model. For external 15-model ensemble predictions, recalibrated HER2 expected calibration error (ECE) was 0.236, exceeding the ECE reference ceiling of 0.125, while calibration-in-the-large was -1.657 [-2.373 to -1.023] (profile-likelihood 95% CI), indicating systematic overprediction. ER and PR also showed overprediction. In paired fold-mean comparisons, HER2 ECE and Brier score favoured the three-transcript model (ESR1, PGR, and ERBB2). Conclusions. A multimodal deep-learning model achieved high discrimination across cohorts but did not demonstrate clear incremental value over a parsimonious three-transcript baseline, while cross-cohort calibration remained a substantial challenge. These findings highlight the importance of evaluating incremental value, calibration, and clinical utility alongside discrimination when assessing AI-based prediction models for medical use.

bioinformatics↗

bioRxiv2025

Peptide molecular glues select between BET paralogues by exploiting allosteric sites and conformational dynamics

Achieving selective target inhibition is critical for minimising drug side effects. This can be especially challenging when targeting individual members of protein families with high sequence similarity. A well-recognised example is the Bromodomain and Extraterminal domain (BET) family of proteins. Chemical inhibition of the acetyllysine (AcK)-binding bromodomains (BDs) of BET proteins has shown considerable promise in a range of disease models. However, despite over a decade of medicinal chemistry efforts, it has proven challenging to develop BET BD inhibitors that exhibit high selectivity between BET family paralogues. Cyclic peptides are versatile scaffolds for therapeutic development and often exhibit high selectivity and affinity for their targets. To explore their potential as selective BET BD inhibitors, we have used mRNA display to identify cyclic peptide ligands for the N-terminal BD of BRD2 and BRD4. The structurally diverse cyclic peptides enriched from the selections boast superior selectivity and affinity to previously developed inhibitors. Most strikingly, we isolated cyclic peptides with [~]1000-fold higher affinity for their target BD over the paralogous BDs, far surpassing selectivities reported to date. Our biochemical and structural data suggest that paralogue-selective cyclic peptides act as molecular glues, exploiting both subtle sequence differences at locations far from the AcK-binding pocket and differences in conformational dynamics between BET BD paralogues to achieve this unprecedented level of specificity. This work provides a blueprint for the development of new classes of selective BET inhibitors and, more generally, underscores the potential of exploiting protein dynamics in the design of selective ligands.

biochemistry↗

bioRxiv2023

The T Cell Receptor beta Chain Repertoire of Tumor Infiltrating Lymphocytes Improves Neoantigen Prediction and Prioritization

In the realm of cancer immunotherapy, the meticulous selection of neoantigens plays a fundamental role in enhancing personalized treatments. Traditionally, this selection process has heavily relied on predicting the binding of peptides to human leukocyte antigens (pHLA). Nevertheless, this approach often overlooks the dynamic interaction between tumor cells and the immune system. In response to this limitation, we have developed an innovative prediction algorithm rooted in machine learning, integrating T cell receptor {beta} chain (TCR{beta}) profiling data from colorectal cancer (CRC) patients for a more precise neoantigen prioritization. TCR{beta} sequencing was conducted to profile the TCR repertoire of tumor-infiltrating lymphocytes (TILs) from 28 CRC patients. The data unveiled both intra-tumor and inter-patient heterogeneity in the TCR{beta} repertoires of CRC patients, likely resulting from the stochastic utilization of V and J segments in response to neoantigens. Our novel combined model integrates pHLA binding information with pHLA-TCR binding to prioritize neoantigens, resulting in heightened specificity and sensitivity compared to models using individual features alone. The efficacy of our proposed model was corroborated through ELISpot assays on long peptides, performed on four CRC patients. These assays demonstrated that neoantigen candidates prioritized by our combined model outperformed predictions made by the established tool NetMHCpan. This comprehensive assessment underscores the significance of integrating pHLA binding with pHLA-TCR binding analysis for more effective immunotherapeutic strategies.

cancer biology↗

bioRxiv2023

Improvement in Neoantigen Prediction via Integration of RNA Sequencing Data for Variant Calling

Neoantigen-based immunotherapy has emerged as a promising strategy for improving the life expectancy of cancer patients. This therapeutic approach heavily relies on accurate identification of cancer mutations using DNA sequencing (DNAseq) data. However, current workflows tend to provide a large number of neoantigen candidates, of which only a limited number elicit efficient and immunogenic T-cell responses suitable for downstream clinical evaluation. To overcome this limitation and increase the number of high-quality immunogenic neoantigens, we propose integrating RNA sequencing (RNAseq) data into the mutation identification step in the neoantigen prediction workflow. In this study, we characterize the mutation profiles identified from DNAseq and/or RNAseq data in tumor tissues of 25 patients with colorectal cancer (CRC). We detected only 22.4% of variants shared between the two methods. In contrast, RNAseq-derived variants displayed unique features of affinity and immunogenicity. We further established that neoantigen candidates identified by RNAseq data significantly increased the number of highly immunogenic neoantigens (confirmed by ELISpot) that would otherwise be overlooked if relying solely on DNAseq data. In conclusion, this integrative approach holds great potential for improving the selection of neoantigens for personalized cancer immunotherapy, ultimately leading to enhanced treatment outcomes and improved survival rates for cancer patients.

immunology↗

bioRxiv2021

DIRECT DETECTION OF LISTERIA MONOCYTOGENES BY RECOMBINASE POLYMERASE AMPLIFICATION

Listeria monocytogenes is one of the most common types of food poisoning bacteria which can cause serious foodborne diseases or even lethality. Generally, L. monocytogenes can be detected using traditional microbiology or molecular biology techniques, notably PCR. However, the application of these methods at the field is restricted due to the strict requirement of equipment and skilled personnel. In this study, recombinase polymerase amplification (RPA), an isothermal PCR assay was developed to rapidly detect L. monocytogenes in the crude samples. The results showed that the RPA reaction, without requiring complex thermal cycles, was well-performed in the optimal conditions of 39{degrees}C within only 25 minutes. The limit of detection was identified as 310 fg of L. monocytogenes genomic DNA, which was 1000-fold more sensitive than the conventional PCR. In addition, RPA also succeeded to directly detect L. monocytogenes cells at a concentration as low as 2.5 x 101 Colony Forming Unit (CFU)/ml in pure cultures and 2.5 x 102 CFU/ml in crude samples without sample extraction or processing. Therefore, RPA established in this study could be an alternative standard method to confirm the presence of L. monocytogenes in food. Accordingly, this rapid and sensitive method could be further applied to clinical testing for the diagnosis of L. monocytogenes infection, especially in areas with limited settings.

molecular biology↗

bioRxiv2020

Direct colorimetric LAMP assay for in-field detection of African swine fever virus: a validation study during an outbreak in Vietnam

African Swine Fever (ASF) is a highly infectious viral disease with high mortality. The most recent ASF outbreak in Vietnam occurred in 2019, posing a threat to spread to the neighboring Asian countries. Without a commercial vaccine or efficient chemotherapeutics successfully developed, rapid diagnosis and necessary biosecurity procedures are required to control the disease. While the diagnosis method of ASF recommended by the World Organization of Animal Health is real-time PCR, it is not suitable for in-field detection of the disease. In this study, a colorimetric Loop-Mediated Isothermal Amplification (LAMP) assay was developed and evaluated for ASF virus detection using crude serum samples collected from domestic pigs in Vietnam during the 2019 outbreak. The LAMP results can be readily visualized to naked eyes within 30 minutes without the requirement of DNA extraction and sophisticated equipment. The sensitivity, specificity, and limit of detection of colorimetric LAMP assay were comparable to a commercial diagnostic real-time PCR kit. Results strongly indicate that the developed colorimetric LAMP assay is highly recommended for the in-field diagnosis of ASF.

molecular biology↗

bioRxiv2020

A comparative study of isothermal nucleic acid amplification methods for SARS-CoV-2 detection at point of care

COVID-19, caused by the novel coronavirus SARS-CoV-2, has spread worldwide and put most of the world under lockdown. Despite that there have been emergently approved vaccines for SARS-CoV-2, COVID-19 cases, hospitalizations, and deaths have remained rising. Thus, rapid diagnosis and necessary public health measures are still key parts to contain the pandemic. In this study, the colorimetric isothermal nucleic acid amplification tests (iNAATs) for SARS-CoV-2 detection based on loop-mediated isothermal amplification (LAMP), cross-priming amplification (CPA), and polymerase spiral reaction (PSR) were designed and evaluated. The three methods showed the same limit of detection (LOD) value of 1 copy of the targeted gene per reaction. However, for the direct detection of SARS-CoV-2 genomic-RNA, LAMP outperformed both CPA and PSR, exhibiting the LOD value of roughly 43.14 genome copies/reaction. The results can be read with the naked eye within 45 minutes, without cross-reactivity to closely related coronaviruses. Moreover, the direct detection of SARS-CoV-2 RNA in simulated patient specimens by iNAATs was also successful. Finally, the ready-to-use lyophilized reagents for LAMP reactions were shown to maintain the sensitivity and LOD value of the liquid assays. The results indicate that the colorimetric lyophilized LAMP kit developed herein is highly suitable for detecting SARS-CoV-2 nucleic acids at point-of-care.

molecular biology↗

bioRxiv2019

Cyclic peptides can engage a single binding pocket through multiple, entirely divergent modes

Cyclic peptide display screening techniques can identify drug leads and biological probes with exceptional affinity and specificity. To date, however, the structural and functional diversity encoded in such peptide libraries remains unexplored. We have used the Random nonstandard Peptide Integrated Discovery (RaPID) system to develop cyclic peptide inhibitors of several acetyllysine-binding bromodomains from the Bromodomain and Extra-Terminal domain (BET) family of epigenetic regulators. These peptides have very high affinities for their targets and exhibit extraordinary selectivity (up to 106-fold), making them the highest-affinity and most specific BET-binding molecules discovered to date. Crystal structures of 13 distinct peptide-bromodomain complexes, which all target the acetyllysine-binding pocket, reveal remarkable diversity in both peptide structure and binding mode, and include both -helical and {beta}-sheet type structures. The peptides can exhibit a high degree of structural pre-organization and bivalent binding of two BDs by one peptide was common, flagging the potential for a new direction in inhibitor design that could bring stronger discrimination between BET-family paralogues. Our data demonstrate for the first time the enormous potential held in these libraries to provide a wide array of modes against a single target, maximizing the opportunity to attain high potency and specificity ligands to a wide variety of proteins.

biochemistry↗