Wang, Xin, et al. "Deep learning using bulk RNA-seq data expands cell landscape identification in tumor microenvironment."?Oncoimmunology?11.1 (2022): 2043662.?https://www.tandfonline.com/doi/full/10.1080/2162402X.2022.2043662
被引次数：5
2022年分区：大类医学2区；小类免疫学2区，肿瘤学 3区
IF 7.723 JCR Q1
一、数据集
从GDC数据库（https://portal.gdc.cancer.gov/projects）下载32种“癌症样本和正常样本”，共有10906患者的表达谱和临床数据。
非小细胞肺癌数据的两种亚型：肺腺癌（LUAD）和肺鳞状细胞癌（LUSC）
从GEO数据集（https://www.ncbi.nlm.nih.gov/geo/）下载5个数据集：
- GSE81861 (CRC): 11 samples of colorectal cancer patients, 7 cell types.Citation32
- GSE75688 (BC): Single-cell sequencing data of 11 breast cancer cells and lymph node metastases.Citation33?It is including 515 single cell RNA-sequencing data, which was sequenced with Illumina HiSeq 2500. The gene expression level is represented as TPM (Transcript per million), which are already normalized and can be comparable across samples in the next analysis.
- GSE86146 (FGC): 2167 individual PGCs and their gonadal niche cells, covering the developmental stages of female and male human embryos from 4 to 26?weeks after pregnancy.Citation21?
- GSE72056 (Melanoma): 31 melanoma samples and their 6 cell types.Citation34?GSE78220 (PD1): Transcriptome samples and corresponding clinical data of 38 melanoma biopsy specimens before anti-PD-1 treatment.Citation35?The cell types and corresponding marker gene are obtained from the CellMarkerCitation36?database (CellMarker2.0).