Study of CITE-seq protein expression prediction method based on LSTM
DOI:
https://doi.org/10.62051/qcng8m62Keywords:
CITE-seq; mRNA sequencing; labeled antibody sequencing; LSTM algorithm; neural network.Abstract
CITE-seq consists of mRNA sequencing and labeled antibody sequencing. This paper predicts labeled antibody sequences based on LSTM neural network. Research shows that LSTM algorithm has a reliable application in CITE-seq protein expression prediction technology, and the prediction accuracy rate reaches 95%. The LSTM algorithm proposed in this paper is relatively accurate and can simulate the gene sequence of proteins well. The CITE-seq data training model learns the potential relationship between RNA and protein, realizes the prediction of protein expression by using scRNA-seq data, greatly reduces the cost of CITE-seq test experiment, and improves the experimental efficiency.
Downloads
References
Musu Y, Liang C, Deng M. Clustering single cell CITE-seq data with a canonical correlation based deep learning method [J]. Cold Spring Harbor Laboratory, 2021. DOI: 10.1101/2021. 09. 07. 459236.
A F X, A S W, A X D, et al. Ensemble learning models that predict surface protein abundance from single-cell multimodal omics data – ScienceDirect [J]. Methods, 2020. DOI: 10.1016/j. ymeth. 2020. 10. 001.
Liu X, Gosline S J, Pflieger L T, et al. Knowledge-based classification of fine-grained immune cell types in single-cell RNA-Seq data with ImmClassifier [J]. Briefings in Bioinformatics, 2021. DOI: 10. 1101/2020. 03. 23. 002758.
Meng X, Yao J. Impact of classification difficulty on the weight matrices spectra in Deep Learning and application to early-stopping [J]. 2021. DOI: 10. 48550/arXiv. 2111. 13331.
Downloads
Published
Conference Proceedings Volume
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
How to Cite
