Dataset Information

RespectM revealed metabolic heterogeneity powers deep learning for reshaping the DBTL cycle

ABSTRACT:

Synthetic biology, relying on Design-Build-Test-Learn (DBTL) cycle, aims to solve medicine, manufacturing and agriculture problems. However, the DBTL cycle’s Learn (L) step lacks predictive power for the behavior of biological systems, resulting from the incompatibility between sparse testing data and chaotic metabolic networks. Herein, we develop a method, 'RespectM', based on mass spectrometry imaging, which is able to detect metabolites at a rate of 500 cells per hour with high efficiency. In this study, 4321 single cell level metabolomics data were acquired, representing metabolic heterogeneity. An optimizable deep neural network was applied to learn from metabolic heterogeneity and a 'heterogeneity-powered learning (HPL)' based model was trained as well. By testing the HPL based model, we suggest minimal operations to achieve high triglyceride production for engineering. The HPL strategy could revolutionize rational design and reshape the DBTL cycle.

INSTRUMENT(S): Liquid Chromatography MS - negative - reverse phase, Liquid Chromatography MS - positive - reverse phase

SUBMITTER: Xuanlin Meng

PROVIDER: MTBLS3830 | MetaboLights | 2023-06-29

REPOSITORIES: MetaboLights

ACCESS DATA

Dataset's files

Source:

			Action	DRS
	MTBLS3830	Other
	FILES	Other
	a_MTBLS3830_LC-MS_negative_reverse-phase_metabolite_profiling.txt	Txt
	a_MTBLS3830_LC-MS_positive_reverse-phase_metabolite_profiling.txt	Txt
	files-all.json	Other

Items per page:

1 - 5 of 9

Similar Datasets

Project description:The cell cycle is known to regulate cell proliferation and cell fate decisions, the underlying mechanism of which is well-studied and conserved across species and cell types. To date, the cell cycle is receiving a renewed importance due to the rapid advancement of single-cell genomics technology. Especially in the analysis of single-cell gene expression, the cell cycle plays a key role in understanding the expression variation across cell states and cell types. The current study was designed to develop effective tools for assessing and predicting the cell cycle in single-cell gene expression data analysis. We collected both Fluorescent Ubiquitination-based Cell Cycle Indicator (FUCCI) – for determining the cell cycle – and single-cell RNA-seq (scRNA-seq) data from each individual using STRT-seq on the Fluidigm C1 platform. The cells were collected from iPSCs derived from 6 genotyped Yoruba cell lines. The experimental design controlled for C1 processing batch effects, as well as individual and gender effects. Using these data, we developed a supervised approach for predicting the cyclical ordering of single cells in the cell cycle using single-cell gene expression data. We used the FUCCI fluorescent intensities to determine a cyclical ordering of the individual cells, and to assign cell time labels for individual cells representing each cell's position in one complete cell cycle. We estimated the cyclical trend of expression levels for each gene based on the FUCCI-derived cell times, and identified candidate sets of cyclical genes for model training. We trained our model in in 5-fold cross-validation and evaluate the trained model in held-out validation samples and external datasets. We compared the prediction results with existing approaches for estimating the cell cycle using single-cell gene expression data: including unsupervised approaches to construct the cyclical ordering of single cells and approaches to categorically assign the cell cycle to individual cells. These results provide a benchmark for assessing the cell cycle in scRNA-seq data analysis and insights into the effects of the cell cycle on gene expression variation in stem cells.

Dataset Information

RespectM revealed metabolic heterogeneity powers deep learning for reshaping the DBTL cycle

Dataset's files

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets