CCLid: A toolkit to authenticate the genotype and stability of cancer cell lines

CCLid (Cancer Cell Line identification) is designed as a toolkit to address the lack of a publicly available resource for genotype-based cell line authentication. We developed this resource to allow for genotype-matching of any given cancer cell line to the 1,204 unique cell lines found in the CCLE dataset, with support to include additional SNP array datasets. Using the B-allele frequencies (BAFs) for all SNPs found in common between the input data and reference datasets, this tool will allow for a genotype matching operation that trains and uses a logistic model to calculate the probability of the best cell line matches. This is followed by a measure of genetic drift between isogenic lines by look for segments of the genome that have significantly different BAF values. This zenodo dataset contains the (sample x probeset) BAF matrix for the CCLE dataset, as well as supporting datasets to allow mapping of SNP probesets and genotype correction between SNP array technologies (i.e. Affymetrix SNP 6.0 and Illumina HumanOmni 2.5M). This also contains all the metadata for cell line identities in CCLE, GDSC, and gCSI as well their corresponding cellosaurus unique identifies.癌症细胞系鉴定工具 (Cancer Cell Line identification，缩写CCLid) 是为填补当前基于基因型的细胞系鉴定公开可用资源的空白而开发的工具包。我们构建该资源，以支持将任意待测癌症细胞系与CCLE数据集收录的1204种独特细胞系进行基因型匹配，同时支持接入额外的SNP芯片数据集。本工具通过比对输入数据与参考数据集共有的全部单核苷酸多态性 (Single Nucleotide Polymorphism, SNP) 位点的B等位基因频率 (B-allele frequencies, BAF)，可执行基因型匹配操作：通过训练并运用逻辑回归模型，计算最优细胞系匹配的概率。随后，该工具可通过检测基因组中存在显著差异BAF值的片段，评估同基因细胞系间的遗传漂变水平。本Zenodo数据集包含CCLE数据集的（样本×探针组）BAF矩阵，以及可实现SNP探针组映射、不同SNP芯片技术间基因型校正的配套数据集，涵盖Affymetrix SNP 6.0与Illumina HumanOmni 2.5M两类芯片技术。本数据集还涵盖CCLE、GDSC与gCSI三大数据库中所有细胞系的身份元数据，以及它们对应的Cellosaurus唯一标识符。