臺灣族群個人之人類白血球抗原與殺手細胞類免疫球蛋白受體基因體組成分析

Abstract

人類基因組中的殺手細胞類免疫球蛋白受體（KIR）基因很重要，但研究起來因為基因的複雜性所以具有挑戰性。 這些基因透過與人類白血球抗原 (HLA) 相互作用來影響自然殺手 (NK) 細胞的活性。 這種相互作用調節免疫反應並影響各種健康狀況，例如自體免疫疾病和器官移植結果。 然而，KIR基因的遺傳結構複雜，且受遺傳因素影響個體間拷貝數（CN）差異較大，使得KIR分型變得複雜。 台灣族群特有的個體層級 KIR-HLA 組合資料缺乏歷史資料，影響了我們的研究方向。

我們的研究利用了各種工具，即 HLA-VBseq、HISAT-genotype、T1K、DRAGEN、Qztype、graphKIR 和 capKIR，來解釋來自各種來源的KIR 和 HLA 基因資料，例如、合成資料、 panel、 全基因體定序（WGS） 、人類泛基因組參考聯盟 (HPRC)資料庫、台灣人體生物資料庫 (TWB)，以及結合爬蟲收集的資料進行分析。 由於每種工具的性能因資料集而異，因此對其功能進行了仔細評估。在 HLA 方面，T1K 和 HISAT-genotype 基因分型因其卓越的準確性和全面的分型功能而成為 HLA 分型性能較好的工具。 在 KIR 方面，graphKIR 最適合 WGS 資料類型，capKIR 適合 panel 資料類型。

使用 graphKIR 工具，我們詳細了解了台灣族群中 KIR 基因變異和不同單倍型的分佈。 對 TWB 的 1,492 個樣本的分析揭示了固有的遺傳異質性，其特徵是 A 單倍型的頻率異常高，其中 AA 單倍型是最常見的(49.7%)，BB(4.62%)，BX(45.7%)。

隨後，我們利用等位基因頻率網路資料庫（AFND）對台灣幾個亞人群的 KIR 基因頻率進行了全面評估。 我們觀察到特定基因的差異，是有關於人口差異和研究方法差異引起的變異的問題，展示了樣本人群內 KIR 基因的複雜動態。例如，我們看 AFND 數據，會發現一些顯著的頻率差異。以 K96 資料庫中的 KIR2DL1 為例。這個基因在台灣 KIR（K96）亞族群中只出現在約 65.6% 的情況下，這一比例顯著低於其他資料庫。這種差異可能源於每個研究中樣本收案或 PCR-SSP primer 設計的差異。這些因素突顯了基因頻率差異如何為我們提供有關 KIR 基因動態性質的重要性。

為了加深我們對 KIR-HLA 交互作用對多種疾病易感性的潛在影響的了解，我們對 TWB 的參與者資料進行了一項關聯性研究。 我們的研究揭示了 KIR-HLA(T1K) 組合與多種表型的關聯，並透過邏輯迴歸分析進一步驗證。 儘管存在年齡和性別等潛在限制因素，但我們的觀察有助於了解 KIR 基因和 KIR-HLA 交互作用對台灣免疫相關疾病易感性的影響。

The human genome''s Killer-cell Immunoglobulin-like Receptor (KIR) genes are significant yet challenging to study due to their gene complexity. These genes modulate the activity of natural killer (NK) cells through interactions with Human Leukocyte Antigens (HLA), impacting various health conditions like autoimmune diseases and organ transplantation outcomes. KIR gene complexity arises from their varied copy number (CN) amongst individuals due to genetic factors, complicating KIR typing. Due to a lack of KIR-HLA-related data for the Taiwanese population, we see an opportunity to focus our research in this direction.

Our research utilized various tools, namely HLA-VBseq, HISAT-genotype, T1K, DRAGEN, Qztype, graphKIR, and capKIR, to interpret KIR and HLA gene data from diverse sources. These sources included synthetic data, panel, whole-genome sequencing (WGS), the Human Pangenome Reference Consortium (HPRC), Taiwan Biobank (TWB) data sets, and data collected through web crawling for analysis. Since every tool''s performance shifts according to the data set employed, we meticulously evaluated their capabilities. T1K and HISAT-genotyping emerged as high-performing tools for HLA due to their superior precision and comprehensive genotyping features. Regarding KIR, graphKIR is most suited for WGS data types, while capKIR is befitting panel data types.

Using graphKIR, we have gained a clear and detailed picture of the variation in KIR genes and the distribution of different haplotypes within the Taiwanese population. Our analysis of 1,492 samples from TWB revealed a significant diversity in the genetic makeup. Specifically, the AA haplotype was most common, seen in 49.7% of the population, followed by BB and BX haplotypes at 4.6% and 45.7%, respectively.

Subsequently, we conducted an exhaustive assessment of KIR gene frequencies within several Taiwanese subpopulations using the Allele Frequency Net Database (AFND). We observed the differential occurrence of specific genes that raised questions about variants due to population differences and methodological disparities, illustrating complex dynamics of KIR genes within sampled populations. For instance, if we look at the AFND data, we find some notable frequency discrepancies. Take the KIR gene 2DL1 in the K96 database as an example. This gene shows up in only about 65.6% of the Taiwanese KIR(K96) subpopulation, a percentage significantly lower than in other databases. This discrepancy could stem from differences in the enrollment of samples or variations in the design of PCR-SSP primers used in each study. These factors accentuate how gene frequency divergence can provide important insights into the dynamic nature of KIR genes.

To deepen our understanding of the potential influence of KIR-HLA interactions on susceptibility to multiple diseases, we conducted an association study on participant data from TWB. Our research uncovered associations of KIR-HLA(T1K) combinations with numerous phenotypes, further confirmed by logistic regression analysis. Despite prevailing potential limiting factors like age and gender, our observations contribute significantly to comprehending the influence of KIR genes and KIR-HLA(T1K) interactions on susceptibility to immune-related diseases in Taiwan.