❝❞
6.19The day I posted the public dumpling to create the"Space & Single Cell Histology Literature Study Group"of the post, along with a couple of reflection questions. Because numbers were kept in check, the final panelists only settled on the9-bitThere are still more than a dozen of you who contacted me and didn't make it on this shuttle, so I'd like to apologize to you."Apologize.". I will let you know if there are other groups in the future. Study Groups fortutorials、videoAnd so on and so forth later on."Publication of Public Dumpling No.", everyone can pay attention to it.
On Sunday, the group conductedNo. 1Share, byTOP Bacteriarespond in singingJefferyThe presenters, respectively:
-
2020 Single-cell combined spatial transcriptome study of the PDAC tumor microenvironment paper -
Science's latest publication on single-cell spatial metabolomics technology
This tweetIt's shared by TOP Bacteria Literaturetext version，videoThe presentation is already inB-stationPosted;The 2nd tweet in this issueIt was shared by Jeffery Literaturetext version, reprinted from his public dumpling:Sangshin Programming Study Room
Let's move on to the literature reading
contexts
This article was published in January 2020 and is the earliest single-cell binding idling article I could find.
The relevant context of the article is 3 main points:
-
scRNA-seq dissociates prior to sequencing, losing spatial information -
Combining in situ hybridization techniques and scRNA-seq can solve the above problems to some extent, but the limitations of ISH are also obvious, and only a small number of genes can be captured -
Back in 2016, spatial transcriptome technology was born, but the limitation was the lack of single-cell resolution
❝This article combines scRNA-seq and spatial transcriptome, two techniques with complementary strengths, to probe the tumor microenvironment of pancreatic ductal adenocarcinoma
❞
in the end
1. Identification of cell subpopulations
The technical route was relatively simple, and the sample size of the early articles was small, with the main content being based on samples made from two patients (paired samples which are scarce)
Subpopulation annotation of single-cell data from two patients was done first, and then the consistency of the results of subpopulation annotation was looked at for both A,B patients.
Still malignant cells were identified by inferring CNV, and further the presence of two groups of cells with significant CNV differences in patient A was seen in the CNV thermogram. The presence of tumor cells in patients A,B was confirmed by immunofluorescence.
2. Analyzing spatial transcriptome data
Firstly, the sections of patients A,B were partitioned according to the tissue characteristics based on the HE staining images, and some gene expressions corresponding to the partitions could also be seen from the spatial transcriptome data.
From the null expression data alone, several clusters of SPOTs can also be seen through a process similar to the analysis of single-cell transcriptomes, and are consistent with the partitioning based on tissue sections
(Multimodal intersection analysis)
In principle, it is similar to enrichment analysis and is not complicated. The heatmap on the right is viewed column by column, the darker the red color, the more enriched this cell type is in this region. With this analysis, the authors found that in patient A, in addition to being enriched for tumor cells in the cancer region, they were also enriched for fibroblasts.
4. Analysis of cellular subclasses
This section analyzed ductal epithelial cells, macrophages, and DC cells using the same routine. The subclasses were first divided into smaller classes, and then the MIA method was used to see how the subclasses were enriched above the REGION.
5. whether different tumor cells co-localize with different other cell types in the cancer region
Two additional sections were taken from patient A. The cancer region was further subdivided among the three sections. co-localization of cancer cluster1 with fibroblasts was seen in all three sections.
6. linking scRNA-seq and idling from the perspective of cell state
Three expression modules were obtained from single-cell transcriptome data using NMF, and the authors focused on the stress-response-related module.
Subsequently, based on the expression of the stress-response module, the CANCER REGION SPOT was divided into two groups, high and low, and the comparison between the two groups continued to identify the highly expressed genes in the HIGH group. Obviously, the stress-response related genes are already differential genes, so why do this? This was done for at least two purposes:
-
Increase in cell number -
Reflects features other than stress-response
The genes identified represented the regional characteristics of those spots with high stress-response, and then MIA analysis was used to see what cells would be enriched in the regions with high stress-response, and the results showed that iCAF (inflammatory fibroblasts) would be enriched in those regions. (Similar conclusions were made above; the conclusions here are further refined.)
The TCGA bulk sample also verifies the correlation between the iCAF signature and the stress-response module
Immunofluorescence also confirmed this result.
The article also uses unpaired melanoma samples to further illustrate the applicability of MIA analysis.
summarize
