DataSheet_5_Using Domain Based Latent Personal Analysis of B Cell Clone Diversity Patterns to Identify Novel Relationships Between the B Cell Clone Populations in Different Tissues.zip

The B cell population is highly diverse and very skewed. It is divided into clones (B cells with a common mother cell). It is thought that each clone represents an initial B cell receptor specificity. A few clones are very abundant, comprised of hundreds or thousands of B cells while the majority have only a few cells per clone. We suggest a novel method - domain-based latent personal analysis (LPA), a method for spectral exploration of entities in a domain, which can be used to find the spectral spread of sub repertoires within a person. LPA defines a domain-based spectral signature for each sub repertoire. LPA signatures consist of the elements, in our case - the clones, that most differentiate the sub repertoire from the person’s abundance of clones. They include both positive elements, which describe overabundant clones, and negative elements that describe missing clones. The signatures can also be used to compare the sub repertoires they represent to each other. Applying LPA to compare the repertoires found in different tissues, we reiterated previous findings that showed that gut and blood tissues have separate repertoires. We further identify a third branch of clonal patterns typical of the lymphatic organs (Spleen, MLN, and bone marrow) separated from the other two categories. We developed a python version of LPA analysis that can easily be applied to compare clonal distributions - https://github.com/ScanLab-ossi/LPA. It could also be easily adapted to study other skewed sequence populations used in the analysis of B cell receptor populations, for instance, k-mers and V gene usage. These analysis types should allow for inter and intra-repertoire comparisons of diversity, which could revolutionize the way we understand repertoire changes and diversity.

B细胞群具有高度多样性且分布极不均衡。其可划分为克隆（clone）——即拥有共同母细胞的B细胞。据推测，每个克隆对应一种初始B细胞受体（B cell receptor）特异性。少数克隆丰度极高，包含数百乃至数千个B细胞，而绝大多数克隆仅含少量细胞。
本研究提出一种全新方法——基于结构域的潜在个人分析（domain-based latent personal analysis, LPA），该方法可用于对某一结构域内的实体进行光谱探索，能够用于识别个体内部子免疫组库（sub repertoire）的光谱分布特征。LPA可为每个子免疫组库定义基于结构域的光谱特征（spectral signature）：此类特征由区分该子免疫组库与个体整体克隆丰度的核心元素（本研究中即为克隆）构成，既包含描述克隆丰度过高的正向元素，也涵盖反映克隆缺失的负向元素。上述特征还可用于对比其所代表的不同子免疫组库。
将LPA应用于不同组织的免疫组库对比分析时，本研究验证了既往研究结论——肠道与血液组织拥有各自独立的免疫组库。此外，本研究进一步发现了一类典型的淋巴器官克隆模式分支，该分支源自脾脏、肠系膜淋巴结（MLN）与骨髓，与前述两类组织的免疫组库存在显著差异。
本研究开发了LPA分析的Python版本工具，可便捷用于克隆分布的对比分析，工具地址为：https://github.com/ScanLab-ossi/LPA。该工具还可轻松适配其他偏倚性序列群体的分析研究，例如B细胞受体群体分析中常用的k聚体（k-mers）与V基因（V gene）使用情况分析。此类分析方法可实现跨免疫组库及免疫组库内部的多样性对比，有望彻底改变我们对免疫组库变化与多样性的认知方式。