Biology subjects

Chiu, Y.-C.

Publications and source records attributed to Chiu, Y.-C..

16 recordsLinked to original sources

bioRxiv2026

spatiAlytica: Viewer-Grounded Multimodal Agentic System for Interactive Spatial Omics Analysis

Spatial transcriptomics and proteomics map tissue architecture and cellular interactions, but analysis remains limited by programming demands and text-centered AI agents that lack viewer grounding and cross-turn context. We present spatiAlytica, a viewer-centric multimodal interactive agentic system embedded in the Napari viewer that enables non-programmer biologists to perform iterative, hypothesis-driven spatial omics analysis via natural language. spatiAlytica couples viewer-state serialization, agentic memory, biological concept-to-data-field mapping, code generation and debugging, Spatial VQA, and grounded interpretation to support an exploratory analysis and interpretive reasoning workflow. We introduce spatiAlyticaBench, a comprehensive benchmark spanning 222 single-turn spatial analytical coding questions, 178 multi-turn sequential workflow questions, and 7,350 image-grounded reasoning questions. spatiAlytica outperformed strong agentic baselines, while using less time and tokens. Case studies across Kaposis sarcoma, colorectal cancer, and ovarian cancer recapitulated known spatial patterns and uncovered progressive CD8 T-cell dysfunction during KS progression.

bioinformatics↗

bioRxiv2026

STiLE: Automated Tissue Microarray Dearraying for Spatial Transcriptomics

Tissue microarrays (TMAs) enable high-throughput spatial transcriptomic profiling of dozens of tissue cores on a single slide. However, existing dearraying methods operate on histological images and do not support the coordinate-based outputs of spatial transcriptomics platforms. Therefore, task of assigning cells to their respective cores (dearraying) remains a manual bottleneck. We present STiLE, a tool for automated TMA dearraying that operates solely on cell centroid coordinates. By eliminating dependence on image data, STiLE is robust to artifacts such as variable staining quality and uneven illumination. The algorithm combines connectivity-based component detection, density-based clustering (HDBSCAN), component-guided cluster merging, and optional grid-based peak detection. Validation on eleven public TMA samples (50-150 cores, three platforms) achieved ARI > 0.99, while systematic benchmarking on 396 synthetic datasets with realistic artifacts demonstrated consistently robust performance (mean ARI = 0.992). STiLE accepts standard formats (AnnData, CSV) and is platform-agnostic, supporting diverse platforms including Vizgen MERSCOPE, 10x Xenium, and NanoString CosMx. An interactive Streamlit interface enables parameter tuning, visual inspection, and region-based processing for large slides. Availability and Implementationhttps://pypi.org/project/stile; source at https://github.com/Huang-AI4Medicine-Lab/stile Contactyu Supplementary InformationSupplementary data are available at Bioinformatics online

bioinformatics↗

bioRxiv2026

Co-repression of Yap1 and Sox9 Abrogates Established Cholangiocarcinoma by Eliminating Transcriptional Compensation

Background/AimsIntrahepatic cholangiocarcinoma (iCCA) represents an unmet clinical need due to its increasing incidence, aggressive biology, and limited treatment options. The extremely low-response rates to current systemic regimens and the emergence of adaptive resistance to targeted therapies underscore the urgent need for alternative therapeutic strategies. Given that the lineage-defining transcription factors SOX9 and YAP1 are central regulators of cholangiocyte and iCCA identity, we investigated their functional roles as potential therapeutic vulnerabilities across multiple preclinical models. MethodsPatient tissue-microarray (TMA) analysis, Sleeping-Beauty hydrodynamic tail vein injection-based iCCA models, and Cre-mediated inducible gene deletion systems were used to investigate the roles of Sox9 and Yap1. Deep-learning-based prediction, RNA-seq, ChIP-seq and immunohistochemistry analyses were performed to delineate transcriptional networks and downstream effectors associated with SOX9/YAP1 signaling. ResultsDual deletion of Sox9 and Yap1 effectively eradicated advanced iCCA while preserving intrahepatic bile ducts, regardless of oncogenic drivers. Mechanistically, SOX9 and YAP1 transcriptionally compensated for each other when one was absent, and ILF2, MGAT5, and WWTR1 were identified as key downstream effectors mediating this compensatory mechanism. Loss of Ilf2, Mgat5, or Taz suppressed iCCA, whereas overexpression of Ilf2 or Taz following Sox9/Yap1 co-deletion restored tumor development, indicating that ILF2 or TAZ can functionally substitute for YAP1 and SOX9 in sustaining iCCA. ConclusionsCo-targeting SOX9 and YAP1 offers a promising and safe broad-spectrum preventive/therapeutic approach for iCCA, potentially overcoming resistance to YAP1 inhibition. The adaptive resistance mechanism identified may extend to other malignancies, providing insights for addressing the advanced resistant to YAP1-TEAD-directed therapies.

cancer biology↗

bioRxiv2025

Cryo-electron tomography reveals lineage-specific replication features of Zika virus

Zika virus (ZIKV), a member of the Flavivirus genus, can cause severe neurodevelopmental disorders when infection occurs during pregnancy. Phylogenetic analyses classify ZIKV into two major lineages: African and Asian. Despite their genetic similarity, only the Asian lineage has been linked to congenital anomalies. ZIKV possesses a single-stranded RNA genome and replicates by remodelling the host endoplasmic reticulum (ER) to form membrane-bound replication complexes, which support viral RNA synthesis, virion assembly, and immune evasion. ZIKV also induces the formation of convoluted membranes (CMs), whose function remains unclear. Most ultrastructural studies of ZIKV replication have relied on transmission electron microscopy, which, despite its value, can alter membrane architecture due to sample preparation. To overcome these limitations, we employed cryo-focused ion beam milling combined with electron tomography to visualize ZIKV replication in a near-native state. By comparing infections with African and Asian ZIKV lineages, we uncovered lineage-specific ultrastructural features, despite similar replication dynamics and growth curves. In cells infected with the Asian strain, virions at different maturation stages were often co-packaged within the same compartment. In contrast, the African lineage ZIKV showed distinct compartmentalization of virions and vesicles. Furthermore, the ER-virion spacing was uniform in African-lineage-infected cells, while morphological heterogeneity was evident in the Asian lineage, including irregular viral sacs. Vesicle diameter was also significantly larger in cells infected with the Asian lineage. Together, these findings reveal both conserved and distinct replication structures between ZIKV lineages. The Asian lineage induces more extensive membrane remodelling and CM formation, potentially contributing to its association with congenital disease.

cell biology↗

bioRxiv2024

Targeting IL-1/IRAK1/4 signaling in Acute Myeloid Leukemia Stem Cells Following Treatment and Relapse

Therapies for acute myeloid leukemia (AML) face formidable challenges due to relapse, often driven by leukemia stem cells (LSCs). Strategies targeting LSCs hold promise for enhancing outcomes, yet paired comparisons of functionally defined LSCs at diagnosis and relapse remain underexplored. We present transcriptome analyses of functionally defined LSC populations at diagnosis and relapse, revealing significant alterations in IL-1 signaling. Interleukin-1 receptor type I (IL1R1) and interleukin-1 receptor accessory protein (IL1RAP) were notably upregulated in leukemia stem and progenitor cells at both diagnosis and relapse. Knockdown of IL1R1 and IL1RAP reduced the clonogenicity and/or engraftment of primary human AML cells. In leukemic MLL-AF9 mice, Il1r1 knockout reduced LSC frequency and extended survival. To target IL-1 signaling at both diagnosis and relapse, we developed UR241-2, a novel interleukin-1 receptor-associated kinase 1 and 4 (IRAK1/4) inhibitor. UR241-2 robustly suppressed IL-1/IRAK1/4 signaling, including NF-{kappa}B activation and phosphorylation of p65 and p38, following IL-1 stimulation. UR241-2 selectively inhibited LSC clonogenicity in primary human AML cells at both diagnosis and relapse, while sparing normal hematopoietic stem and progenitor cells. It also reduced AML engraftment in leukemic mice. Our findings highlight the therapeutic potential of UR241-2 in targeting IL-1/IRAK1/4 signaling to eradicate LSCs and improve AML outcomes.

cancer biology↗

bioRxiv2024

Spatial Transcriptomic Profiling Of Advanced Ovarian Clear Cell Carcinoma Reveals Intra-Tumor Heterogeneity In Epithelial-Mesenchymal Gradient

Intratumoral heterogeneity is intrinsically comprised of molecular alterations of tumor cells and extrinsically from interconnections with microenvironments. This study explores the spatial heterogeneity of ovarian clear cell carcinoma (OCCC), a rare cancer with significance to East Asian women. We profile 21 primary-metastatic tumor pairs in a discovery cohort and 16 tumors in two validation cohorts using spatial transcriptomic (ST) platforms. Our integrative analysis revealed an inverse relationship between OXPHOS and inflammation along the EMT gradient. OCCC cells undergoing partial EMT have metabolic shifts and lose LCN2 expression, possibly via concomitant down-regulation of SOX9. Conversely, LCN2 expression correlated with OXPHOS-enriched tumor signature, low EMT, and better outcomes in OCCC. Single-cell ST profiling using CosMx further identifies nine spatially distinct cancer cell populations including the LCN2-high cancer subclone with a high epithelial score. SOX9 induction could partially restore epithelial-ness in LCN2-low cells suggesting that plasticity in OCCC is achieved via transcriptional reprogramming. Our findings provide further insights into epithelial-mesenchymal plasticity and the adaptive interactions between cancer cells and their microenvironments in OCCC.

cancer biology↗

bioRxiv2024

High-Throughput Empirical and Virtual Screening to Discover Novel Inhibitors of Polyploid Giant Cancer Cells in Breast Cancer

Therapy resistance in breast cancer is increasingly attributed to polyploid giant cancer cells (PGCCs), which arise through whole-genome doubling and exhibit heightened resilience to standard treatments. Characterized by enlarged nuclei and increased DNA content, these cells tend to be dormant under therapeutic stress, driving disease relapse. Despite their critical role in resistance, strategies to effectively target PGCCs are limited, largely due to the lack of high-throughput methods for assessing their viability. Traditional assays lack the sensitivity needed to detect PGCC-specific elimination, prompting the development of novel approaches. To address this challenge, we developed a high-throughput single-cell morphological analysis workflow designed to differentiate compounds that selectively inhibit non-PGCCs, PGCCs, or both. Using this method, we screened a library of 2,726 FDA Phase 1-approved drugs, identifying promising anti-PGCC candidates, including proteasome inhibitors, FOXM1, CHK, and macrocyclic lactones. Notably, RNA-Seq analysis of cells treated with the macrocyclic lactone Pyronaridine revealed AXL inhibition as a potential strategy for targeting PGCCs. Although our single-cell morphological analysis pipeline is powerful, empirically testing all existing compounds is impractical and inefficient. To overcome this limitation, we trained a machine learning model to predict anti-PGCC efficacy in silico, integrating chemical fingerprints and compound descriptions from prior publications and databases. The model demonstrated a high correlation with experimental outcomes and predicted efficacious compounds in an expanded library of over 6,000 drugs. Among the top-ranked predictions, we experimentally validated two compounds as potent PGCC inhibitors. These findings underscore the synergistic potential of integrating high-throughput empirical screening with machine learning-based virtual screening to accelerate the discovery of novel therapies, particularly for targeting therapy-resistant PGCCs in breast cancer.

bioengineering↗

bioRxiv2024

reguloGPT: Harnessing GPT for Knowledge Graph Construction of Molecular Regulatory Pathways

MotivationMolecular Regulatory Pathways (MRPs) are crucial for understanding biological functions. Knowledge Graphs (KGs) have become vital in organizing and analyzing MRPs, providing structured representations of complex biological interactions. Current tools for mining KGs from biomedical literature are inadequate in capturing complex, hierarchical relationships and contextual information about MRPs. Large Language Models (LLMs) like GPT-4 offer a promising solution, with advanced capabilities to decipher the intricate nuances of language. However, their potential for end-to-end KG construction, particularly for MRPs, remains largely unexplored. ResultsWe present reguloGPT, a novel GPT-4 based in-context learning prompt, designed for the end-to-end joint name entity recognition, N-ary relationship extraction, and context predictions from a sentence that describes regulatory interactions with MRPs. Our reguloGPT approach introduces a context-aware relational graph that effectively embodies the hierarchical structure of MRPs and resolves semantic inconsistencies by embedding context directly within relational edges. We created a benchmark dataset including 400 annotated PubMed titles on N6-methyladenosine (m6A) regulations. Rigorous evaluation of reguloGPT on the benchmark dataset demonstrated marked improvement over existing algorithms. We further developed a novel G-Eval scheme, leveraging GPT-4 for annotation-free performance evaluation and demonstrated its agreement with traditional annotation-based evaluations. Utilizing reguloGPT predictions on m6A-related titles, we constructed the m6A-KG and demonstrated its utility in elucidating m6As regulatory mechanisms in cancer phenotypes across various cancers. These results underscore reguloGPTs transformative potential for extracting biological knowledge from the literature. Availability and implementationThe source code of reguloGPT, the m6A title and benchmark datasets, and m6A-KG are available at: https://github.com/Huang-AI4Medicine-Lab/reguloGPT.

bioinformatics↗

bioRxiv2023

Mpox virus spreads from cell-to-cell and leads to neuronal injury in human cerebral organoids

In 2022-23, the world experienced the largest recorded monkeypox virus (MPXV) outbreak outside of endemic regions. Remarkably, cases of neurological manifestations were reported, some of which fatal. MPXV DNA and MPXV-specific antibodies were detected in the cerebrospinal fluid of encephalitis-affected patients, suggesting neuroinvasive potential of MPXV. We explored the susceptibility of neural tissue to MPXV infection using human neural organoids (hNOs) exposed to a primary isolate belonging to clade IIb lineage. The virus efficiently replicates in hNOs as indicated by the exponential increase of infectious viral loads and the elevated frequency of MPXV-positive cells over time. Electron microscopy imaging revealed the presence of viral particles as well as perinuclear viral factories. We observed susceptibility of several cell types to the virus, including neural progenitor cells and neurons. Furthermore, we detected the presence of viral antigen in neurites and in foci of grouped cells distributed throughout the tissue. In line with this, we documented significantly more cell-associated than released infectious virus, suggesting viral spread by cell-to-cell contact. Using an mNeonGreen-expressing recombinant MPXV, we confirmed cell-associated virus transmission through live-cell imaging. While hNOs displayed no evident outer morphological changes upon infection, we detected the formation of beads in neurites, a phenomenon commonly associated with neurodegenerative disorders. Live-cell imaging further confirmed the recurrent formation of neuritic beads in neurons in the days following MPXV infection, with bead formation preceding neurite-initiated cell death. Notably, treatment of MPXV infected hNOs with the antiviral drug tecovirimat resulted in a significant reduction of infectious viral loads by several orders of magnitude. Taken together, our findings suggest viral manipulation of axonal transport driving neuronal degeneration and identify a mechanism potentially contributing to MPXV-mediated neuropathology that may have therapeutic implications.

microbiology↗

bioRxiv2022

DepLink: an R Shiny app to systematically link genetic and pharmacologic dependencies of cancer

Large-scale genetic and pharmacologic dependency maps are generated to reveal genetic vulnerabilities and drug sensitivities of cancer. However, user-friendly software is needed to systematically link such maps. Here we present DepLink, an R Shiny server to identify genetic and pharmacologic perturbations that induce similar effects on cell viability or molecular changes. DepLink integrates heterogeneous datasets of genome-wide CRISPR loss-of-function screens, high-throughput pharmacologic screens, and perturbation expression signatures. The datasets are systematically connected by four complementary modules tailored for different query scenarios. In summary, DepLink enables easy navigation, visualization, and linkage of rapidly evolving cancer dependency maps.

bioinformatics↗

bioRxiv2021

Comparative Genome Analysis of 'Candidatus Phytoplasma luffae' Reveals the Influential Roles of Potential Mobile Units in Phytoplasma Evolution

Phytoplasmas are insect-transmitted plant pathogens that cause substantial losses in agriculture. In addition to economic impact, phytoplasmas induce distinct disease symptoms in infected plants, thus attracting attention for research on molecular plant-microbe interactions and plant developmental processes. Due to the difficulty of establishing an axenic culture of these bacteria, culture-independent genome characterization is a crucial tool for phytoplasma research. However, phytoplasma genomes have strong nucleotide composition biases and are repetitive, which make it challenging to produce complete assemblies. In this study, we utilized Illumina and Oxford Nanopore sequencing technologies to obtain the complete genome sequence of Candidatus Phytoplasma luffae strain NCHU2019 that is associated with witches broom disease of loofah (Luffa aegyptiaca) in Taiwan. The fully assembled circular chromosome is 769 kb in size and is the first representative genome sequence of group 16SrVIII phytoplasmas. Comparative analysis with other phytoplasmas revealed that NCHU2019 has an exceptionally repetitive genome, possessing a pair of 75 kb repeats and at least 13 potential mobile units (PMUs) that account for ~25% of its chromosome. This level of genome repetitiveness is exceptional for bacteria, particularly among obligate pathogens with reduced genomes. Our genus-level analysis of PMUs demonstrated that these phytoplasma-specific mobile genetic elements can be classified into three major types that differ in gene organization and phylogenetic distribution. Notably, PMU abundance explains nearly 80% of the variance in phytoplasma genome sizes, a finding that provides a quantitative estimate for the importance of PMUs in phytoplasma genome variability. Finally, our investigation found that in addition to horizontal gene transfer, PMUs also contribute to intra-genomic duplications of effector genes, which may provide redundancy for neofunctionalization or subfunctionalization. Taken together, this work improves the taxon sampling for phytoplasma genome research and provides novel information regarding the roles of mobile genetic elements in phytoplasma evolution.

genomics↗

bioRxiv2021

Teosinte introgression modulates phosphatidylcholine levels and induces early maize flowering time

Native Americans domesticated maize (Zea mays ssp. mays) from lowland teosinte parviglumis (Zea mays ssp.parviglumis) in the warm Mexican southwest and brought it to the highlands of Mexico and South America where it was exposed to lower temperatures that imposed strong selection on flowering time. Phospholipids are important metabolites in plant responses to low-temperature and phosphorus availability, and have been suggested to influence flowering time. Here, we combined linkage mapping with genome scans to identify High PhosphatidylCholine 1 (HPC1), a gene that encodes a phospholipase A1 enzyme, as a major driver of phospholipid variation in highland maize. Common garden experiments demonstrated strong genotype-by-environment interactions associated with variation at HPC1, with the highland HPC1 allele leading to higher fitness in highlands, possibly by hastening flowering. The highland maize HPC1 variant resulted in impaired function of the encoded protein due to a polymorphism in a highly conserved sequence. A meta-analysis across HPC1 orthologs indicated a strong association between the identity of the amino acid at this position and optimal growth in prokaryotes. Mutagenesis of HPC1 via genome editing validated its role in regulating phospholipid metabolism. Finally, we showed that the highland HPC1 allele entered cultivated maize by introgression from the wild highland teosinte Zea mays ssp. mexicana and has been maintained in maize breeding lines from the Northern US, Canada and Europe. Thus, HPC1 introgressed from teosinte mexicana underlies a large metabolic QTL that modulates phosphatidylcholine levels and has an adaptive effect at least in part via induction of early flowering time.

plant biology↗

bioRxiv2020

CancerSiamese: one-shot learning for primary and metastatic tumor classification

We consider cancer classification based on one single gene expression profile. We proposed CancerSiamese, a new one-shot learning model, to predict the cancer type of a query primary or metastatic tumor sample based on a support set that contains only one known sample for each cancer type. CancerSiamese receives pairs of gene expression profiles and learns a representation of similar or dissimilar cancer types through two parallel Convolutional Neural Networks joined by a similarity function. We trained CancerSiamese for both primary and metastatic cancer type predictions using samples from TCGA and MET500. Test results for different N-way predictions yielded an average accuracy improvement of 8% and 4% over the benchmark 1-Nearest Neighbor (1-NN) classifier for primary and metastatic tumors, respectively. Moreover, we applied the guided gradient saliency map and feature selection to CancerSiamese to identify and analyze the marker-gene candidates for primary and metastatic cancers. Our work demonstrated, for the first time, the feasibility of applying one-shot learning for expression-based cancer type prediction when gene expression data of cancer types are limited and could inspire new and ingenious applications of one-shot and few-shot learning solutions for improving cancer diagnosis, treatment planning, and our understanding of cancer.

bioinformatics↗

bioRxiv2020

Predicting sites of epitranscriptome modifications using unsupervised representation learning based on generative adversarial networks

Epitranscriptome is an exciting area that studies different types of modifications in transcripts and the prediction of such modification sites from the transcript sequence is of significant interest. However, the scarcity of positive sites for most modifications imposes critical challenges for training robust algorithms. To circumvent this problem, we propose MR-GAN, a generative adversarial network (GAN) based model, which is trained in an unsupervised fashion on the entire pre-mRNA sequences to learn a low dimensional embedding of transcriptomic sequences. MR-GAN was then applied to extract embeddings of the sequences in a training dataset we created for eight epitranscriptome modifications, including m6A, m1A, m1G, m2G, m5C, m5U, 2'-O-Me, Pseudouridine ({Psi}) and Dihydrouridine (D), of which the positive samples are very limited. Prediction models were trained based on the embeddings extracted by MR-GAN. We compared the prediction performance with the one-hot encoding of the training sequences and SRAMP, a state-of-the-art m6A site prediction algorithm and demonstrated that the learned embeddings outperform one-hot encoding by a significant margin for up to 15% improvement. Using MR-GAN, we also investigated the sequence motifs for each modification type and uncovered known motifs as well as new motifs not possible with sequences directly. The results demonstrated that transcriptome features extracted using unsupervised learning could lead to high precision for predicting multiple types of epitranscriptome modifications, even when the data size is small and extremely imbalanced.

bioinformatics↗

bioRxiv2020

Assessing the heterogeneity of cardiac non-myocytes and the effect of cell culture with integrative single cell analysis

Cardiac non-myocytes comprise a diverse and crucial cell population in the heart that plays dynamic roles in cardiac wound healing and growth. Non-myocytes broadly fall into four cell types: endothelium, fibroblasts, leukocytes, and pericytes. Here we characterize the diversity of the non-myocytes in vivo and in vitro using mass cytometry. By leveraging single-cell RNA sequencing we inform the design of a mass cytometry panel. To aid in annotation of the mass cytometry datasets, we utilize data integration with a neural network. We introduce approximately 460,000[~] single cell proteomes of non-myocytes as well as 5,000[~] CD31 negative single cell transcriptomes. Using our data, as well as previously reported datasets, we characterize cardiac non-myocytes with high depth in six mice, characterizing novel surface markers (CD9, CD200, Notch3, and FolR2). Further, we find that extended cell culture promotes the proliferation of CD45+CD11b+FolR2+IAIE- myeloid cells in addition to fibroblasts.

cell biology↗

bioRxiv2019

An ancestral informative marker panel design for individual ancestry estimation of Hispanic population using whole exome sequencing data

BackgroundEuropeans and American Indians were major genetic ancestry of Hispanics in the U.S. In those ancestral groups, it has markedly different incidence rates and outcomes in many types of cancers. Therefore, the genetic admixture may cause biased genetic association study with cancer susceptibility variants specifically in Hispanics. The incidence rate and genetic mutational pattern of liver cancer have been shown substantial disparity between Hispanic, Asian and non-Hispanic white populations. Currently, ancestry informative marker (AIM) panels have been widely utilized with up to a few hundred ancestry-informative single nucleotide polymorphisms (SNPs) to infer ancestry admixture. Notably, current available AIMs are predominantly located in intron and intergenic regions, while the whole exome sequencing (WES) protocols commonly used in translational research and clinical practice do not contain these markers, thus, the challenge to accurately determine a patients admixture proportion without subject to additional DNA testing.\n\nMethodsHere we designed a bioinformatics pipeline to obtain an AIM panel. The panel infers 3-way genetic admixture from three distinct continental populations (African (AFR), European (EUR), and East Asian (EAS)) constraint within evolutionary-conserved exome regions. Briefly, we extract [~]1 million exonic SNPs from all individuals of three populations in the 1000 Genomes Project. Then, the SNPs were trimmed by their linkage disequilibrium (LD), restricted to biallelic variants only, and assembled as an AIM panel with the top ancestral informativeness statistics based on the In-statistic. The selected AIM panel was applied to training dataset and clinical dataset. Finally, The ancestral proportions of each individual was estimated by STRUCTURE.\n\nResultsIn this study, the optimally selected AIM panel with 250 markers, or the UT-AIM250 panel, was performed with better accuracy as one of the published AIM panels when we tested with 3 ancestral populations (Accuracy: 0.995 {+/-} 0.012 for AFR, 0.997 {+/-} 0.007 for EUR, and 0.994 {+/-} 0.012 for EAS). We demonstrated the utility of UT-AIM250 panel on the admixed American (AMR) of the 1000 Genomes Project and obtained similar results (AFR: 0.085 {+/-} 0.098; EUR: 0.665 {+/-} 0.182; and EAS 0.250 {+/-} 0.205) to previously published AIM panels (Phillips-AIM34: AFR: 0.096 {+/-} 0.127, EUR: 0.575 {+/-} 0.29; and EAS: 0.330 {+/-} 0.315; Wei-AIM278: AFR: 0.070 {+/-} 0.096, EUR: 0.537 {+/-} 0.267, and EAS: 0.393 {+/-} 0.300) with no significant difference (Pearson correlation, P < 10-50, n = 347 samples). Subsequently, we applied UT-AIM250 panel to clinical datasets of self-reported Hispanic patients in South Texas with hepatocellular carcinoma (26 patients). Our estimated admixture proportions from adjacent non-cancer liver tissue data of Hispanics in South Texas is (AFR: 0.065 {+/-} 0.043; EUR: 0.594 {+/-} 0.150; and EAS: 0.341 {+/-} 0.160), with smaller variation due to its unique Texan/Mexican American population in South Texas. Similar admixture proportion from the corresponding tumor tissue we also obtained. In addition, we estimated admixture proportions of entire TCGA-LIHC samples (376 patients) using UT-AIM250 panel. We demonstrated that our AIM panel estimate consistent admixture proportions from DNAs derived from tumor and normal tissues, and 2 possible incorrect reported race/ethnicity, and/or provide race/ethnicity determination if necessary.\n\nConclusionsTaken together, we demonstrated the feasibility of using evolutionary-conserved exome regions to distinguish genetic ancestry descendants based on 3 continental-ancestry proportion, provided a robust and reliable control for sample collection or patient stratification for genetic analysis. R implementation of UT-AIM250 is available at https://github.com/chenlabgccri/UT-AIM250.

bioinformatics↗