CaSee: A lightning transfer-learning model directly used to discriminate cancer/normal cells from scRNA-seq

Sh, Yuan; Zhang, Xiuli; Yang, Zhimin; Dong, Jierong; Wang, Yuanzhuo; Zhou, Ying; Li, Xuejie; Guo, Caixia; Hu, Zhiyuan

doi:10.1038/s41388-022-02478-5

Cited by 5 publications

(7 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Innovative modifications in scRNA-seq such as Seq-Well S3 and FLASH-seq enable faster library preparation as well as increased sensitivity to genes with low expression profiles [21,22 ▪ ]. Bioinformatic tools such as SMURF enhance scRNA-seq sensitivity by improving signal recovery while CaSee differentiates cancer cells from noncancerous cells allowing more accurate downstream analysis [23,24]. Among classical technologies, qRT-PCR is still reliable and widely used, particularly of late to identify recombination-specific gene expression signatures in solid biopsies [25].…”

Section: Technologies At Specific Omic Levels: Revealing Exploitable ...mentioning

confidence: 99%

Explore & actuate: the future of personalized medicine in oncology through emerging technologies

Babu,

Sen

2024

Current Opinion in Oncology

View full text Add to dashboard Cite

Purpose of review The future of medicine is aimed to equip the physician with tools to assess the individual health of the patient for the uniqueness of the disease that separates it from the rest. The integration of omics technologies into clinical practice, reviewed here, would open new avenues for addressing the spatial and temporal heterogeneity of cancer. The rising cancer burden patiently awaits the advent of such an approach to personalized medicine for routine clinical settings. Recent findings To weigh the translational potential, multiple technologies were categorized based on the extractable information from the different types of samples used, to the various omic-levels of molecular information that each technology has been able to advance over the last 2 years. This review uses a multifaceted classification that helps to assess translational potential in a meaningful way toward clinical adaptation. Summary The importance of distinguishing technologies based on the flow of information from exploration to actuation puts forth a framework that allows the clinicians to better adapt a chosen technology or use them in combination to enhance their goals toward personalized medicine.

show abstract

Section: Technologies At Specific Omic Levels: Revealing Exploitable ...mentioning

confidence: 99%

Explore & actuate: the future of personalized medicine in oncology through emerging technologies

Babu,

Sen

2024

Current Opinion in Oncology

View full text Add to dashboard Cite

show abstract

“…Notably, the regulatory interplay of neuron-related signal factors and hormones within the TME further fosters immune suppression 19 and angiogenesis 20 in breast cancer. Single-cell multiomics sequencing technology has emerged as a valuable tool for investigating intra-tumor heterogeneity and unraveling the complexities of TME across various malignant tumors 21,22 . This approach holds promise for identifying potential therapeutic targets that align with precision medicine principles.…”

Section: Mainmentioning

confidence: 99%

scRNA-seq and spatial transcriptomics reveal neuroendocrine-like cancer cells promote angiogenesis and EMT through neural signaling pathways in male breast cancer

Zhang,

Sh,

Cai

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Male breast cancer (MBC) is a relatively rare and inadequately researched disease, and its cellular and molecular traits remain obscure. In this study, we conducted single-cell sequencing (N=20) and spatial transcriptomics (N=14) on 34 fresh tissue samples from 27 MBC patients. We identified six major cancer cell subtypes that are associated with the development and progression of MBC. Specifically, cancer cells exhibiting neuroendocrine-like properties facilitate immune evasion, tumor angiogenesis, epithelial-to-mesenchymal transition, cell proliferation, tumor invasion, and metastasis. They do so by secreting neuro-related factors and engaging in regulating neuro-related signaling pathways, synergistically interacting with T cells, macrophages, and fibroblasts within the tumor microenvironment. Additionally, we found that mutations or copy number variations amplifications of the UTY gene on the Y chromosome and/or its high transcript expression are closely associated with adverse clinical outcomes in male cancer patients, including MBC patients. In conclusion, our study provides important data support for a deeper understanding of the molecular characteristics and tumor microenvironment of MBC, and offers important clues for developing improved therapeutic strategies to improve the prognosis of MBC patients.

show abstract

“…InferCNV relies on the use of normal cells/spots as a reference and the user’s subjective judgment to discriminate malignant cells/spots. On the other hand, CopyKAT must statistically distinguish between intact and aneuploid, making it difficult to obtain high accuracy with high-purity data (a set of nearly exclusively malignant or non-malignant cells) 14 . In addition, current evidence suggests that cell copy number alterations are widespread in normal human tissues 15 , 16 , which can lead to misclassification.…”

Section: Introductionmentioning

confidence: 99%

“…Machine learning, especially neural networks, has introduced additional concepts for automatic cell/spot annotation. In recent years, among the methods for automatic annotation of malignant state based on machine learning, two representative methods for distinguishing the degree of cell malignancy are ikarus 17 and Casee 14 . Ikarus is based on a logistic regression model.…”

Section: Introductionmentioning

confidence: 99%

Domain generalization enables general cancer cell annotation in single-cell and spatial transcriptomics

Zhong,

Hou,

Yao

et al. 2024

Nat Commun

View full text Add to dashboard Cite

Single-cell and spatial transcriptome sequencing, two recently optimized transcriptome sequencing methods, are increasingly used to study cancer and related diseases. Cell annotation, particularly for malignant cell annotation, is essential and crucial for in-depth analyses in these studies. However, current algorithms lack accuracy and generalization, making it difficult to consistently and rapidly infer malignant cells from pan-cancer data. To address this issue, we present Cancer-Finder, a domain generalization-based deep-learning algorithm that can rapidly identify malignant cells in single-cell data with an average accuracy of 95.16%. More importantly, by replacing the single-cell training data with spatial transcriptomic datasets, Cancer-Finder can accurately identify malignant spots on spatial slides. Applying Cancer-Finder to 5 clear cell renal cell carcinoma spatial transcriptomic samples, Cancer-Finder demonstrates a good ability to identify malignant spots and identifies a gene signature consisting of 10 genes that are significantly co-localized and enriched at the tumor-normal interface and have a strong correlation with the prognosis of clear cell renal cell carcinoma patients. In conclusion, Cancer-Finder is an efficient and extensible tool for malignant cell annotation.

show abstract

CaSee: A lightning transfer-learning model directly used to discriminate cancer/normal cells from scRNA-seq

Cited by 5 publications

References 36 publications

Explore & actuate: the future of personalized medicine in oncology through emerging technologies

Explore & actuate: the future of personalized medicine in oncology through emerging technologies

scRNA-seq and spatial transcriptomics reveal neuroendocrine-like cancer cells promote angiogenesis and EMT through neural signaling pathways in male breast cancer

Domain generalization enables general cancer cell annotation in single-cell and spatial transcriptomics

Contact Info

Product

Resources

About