-
Notifications
You must be signed in to change notification settings - Fork 31
/
Copy pathREADME.Rmd
294 lines (223 loc) · 24.3 KB
/
README.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
---
output: github_document
---
<!-- README.md is generated from README.Rmd. Please edit that file -->
```{r 'setup', include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>",
fig.path = "man/figures/README-",
out.width = "100%"
)
library('BiocStyle')
library('spatialLIBD')
## Bib setup
library('knitcitations')
## Load knitcitations with a clean bibliography
cleanbib()
cite_options(hyperlink = 'to.doc', citation_format = 'text', style = 'html')
# Note links won't show for now due to the following issue
# https://github.com/cboettig/knitcitations/issues/63
## Write bibliography information
bib <- c(
R = citation(),
AnnotationHub = citation('AnnotationHub'),
BiocStyle = citation('BiocStyle'),
cowplot = citation('cowplot'),
DT = citation('DT'),
ExperimentHub = citation('ExperimentHub'),
fields = citation('fields'),
ggplot2 = citation('ggplot2'),
golem = citation('golem'),
IRanges = citation('IRanges'),
knitcitations = citation('knitcitations'),
knitr = citation('knitr')[3],
plotly = citation('plotly'),
Polychrome = citation('Polychrome'),
RColorBrewer = citation('RColorBrewer'),
rmarkdown = citation('rmarkdown')[1],
S4Vectors = citation('S4Vectors'),
scater = citation('scater'),
sessioninfo = citation('sessioninfo'),
SingleCellExperiment = citation('SingleCellExperiment'),
shiny = citation('shiny'),
SummarizedExperiment = citation('SummarizedExperiment'),
testthat = citation('testthat'),
viridisLite = citation('viridisLite'),
spatialLIBD = citation('spatialLIBD')[1],
spatialLIBDpaper = citation('spatialLIBD')[2]
)
```
# HumanPilot <img src="http://research.libd.org/spatialLIBD/reference/figures/logo.png" align="right" />
<!-- badges: start -->
[](https://zenodo.org/badge/latestdoi/225910046)
<!-- badges: end -->
Welcome to the `spatialLIBD` project! It is composed of the `HumanPilot` described here as well as:
* a [shiny](https://shiny.rstudio.com/) web application that we are hosting at [spatial.libd.org/spatialLIBD/](http://spatial.libd.org/spatialLIBD/) that can handle a [limited](https://github.com/LieberInstitute/spatialLIBD/issues/2) set of concurrent users,
* a Bioconductor package at [bioconductor.org/packages/spatialLIBD](http://bioconductor.org/packages/spatialLIBD) (or from [here](http://research.libd.org/spatialLIBD/)) that lets you analyze the data and run a local version of our web application (with our data or yours),
* and a [research article](https://doi.org/10.1038/s41593-020-00787-0) with the scientific knowledge we drew from this dataset. The analysis code for our project is available [here](https://github.com/LieberInstitute/HumanPilot/) that you are looking at right now. The high quality figures for the manuscript are available through [Figshare](https://doi.org/10.6084/m9.figshare.13623902.v1).
The web application allows you to browse the LIBD human dorsolateral pre-frontal cortex (DLPFC) spatial transcriptomics data generated with the 10x Genomics Visium platform. Through the [R/Bioconductor package](https://bioconductor.org/packages/spatialLIBD) you can also download the data as well as visualize your own datasets using this web application. Please check the [manuscript](https://doi.org/10.1038/s41593-020-00787-0) or [bioRxiv pre-print](https://www.biorxiv.org/content/10.1101/2020.02.28.969931v1) for more details about this project.
If you tweet about this website, the data or the R package please use the <code>#spatialLIBD</code> hashtag. You can find previous tweets that way as shown <a href="https://twitter.com/search?q=%23spatialLIBD&src=typed_query">here</a>. Thank you! <a href="https://twitter.com/intent/tweet?button_hashtag=spatialLIBD&ref_src=twsrc%5Etfw" class="twitter-hashtag-button" data-show-count="false">Tweet #spatialLIBD</a><script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
## Study design
As a quick overview, the data presented here is from portion of the
DLPFC that spans six neuronal layers plus white matter (**A**) for a
total of three subjects with two pairs of spatially adjacent replicates
(**B**). Each dissection of DLPFC was designed to span all six layers
plus white matter (**C**). Using this web application you can explore
the expression of known genes such as *SNAP25* (**D**, a neuronal gene),
*MOBP* (**E**, an oligodendrocyte gene), and known layer markers from
mouse studies such as *PCP4* (**F**, a known layer 5 marker gene).
<img src="http://research.libd.org/spatialLIBD/reference/figures/paper_figure1.jpg" align="center" width="800px" />
This web application was built such that we could annotate the spots to
layers as you can see under the **spot-level data** tab. Once we
annotated each spot to a layer, we compressed the information by a
pseudo-bulking approach into **layer-level data**. We then analyzed the
expression through a set of models whose results you can also explore
through this web application. Finally, you can upload your own gene sets
of interest as well as layer enrichment statistics and compare them with
our LIBD Human DLPFC Visium dataset.
If you are interested in running this web application locally, you can
do so thanks to the `spatialLIBD` R/Bioconductor package that powers
this web application as shown below.
```{r run_app, eval = FALSE}
## Run this web application locally
spatialLIBD::run_app()
## You will have more control about the length of the
## session and memory usage.
## You could also use this function to visualize your
## own data given some requirements described
## in detail in the package vignette documentation
## at http://research.libd.org/spatialLIBD/.
```
## Shiny website mirrors
* [Main shiny application website](http://spatial.libd.org/spatialLIBD/) (note that the link must have a trailing slash `/` for it to work)
* [Shinyapps](https://libd.shinyapps.io/spatialLIBD/) This version has less RAM memory but is typically deployed using the latest version of `spatialLIBD`.
## Introductory material
If you prefer to watch a video overview of the `HumanPilot` project, check the following journal club presentation of the main results.
<iframe width="560" height="315" src="https://www.youtube.com/embed/qloLbG5-IPM?si=1gO1fujrgSXPfa6F" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen data-external="1"></iframe>
You might also be interested in the explainer video and [companion blog post](https://lcolladotor.github.io/2024/05/23/humanpilot-first-spatially-resolved-transcriptomics-study-using-visium/) as well as [the original Feb 29, 2020 blog post](https://lcolladotor.github.io/2020/02/29/diving-together-into-the-unknown-world-of-spatial-transcriptomics/) from when we first made this project public.
<iframe width="560" height="315" src="https://www.youtube.com/embed/HGioWKuI3ek?si=X-tqtZtcPSV-3uMt" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen data-external="1"></iframe>
## R/Bioconductor package
The `spatialLIBD` package contains functions for:
- Accessing the spatial transcriptomics data from the LIBD Human Pilot
project ([code on
GitHub](https://github.com/LieberInstitute/HumanPilot)) generated
with the Visium platform from 10x Genomics. The data is retrieved
from [Bioconductor](http://bioconductor.org/)’s `ExperimentHub`.
- Visualizing the spot-level spatial gene expression data and
clusters.
- Inspecting the data interactively either on your computer or through
[spatial.libd.org/spatialLIBD/](http://spatial.libd.org/spatialLIBD/).
For more details, please check the [documentation
website](http://lieberinstitute.github.io/spatialLIBD) or the
Bioconductor package landing page
[here](https://bioconductor.org/packages/spatialLIBD).
## Installation instructions
Get the latest stable `R` release from
[CRAN](http://cran.r-project.org/). Then install `spatialLIBD` using
from [Bioconductor](http://bioconductor.org/) the following code:
```{r 'install', eval = FALSE}
if (!requireNamespace("BiocManager", quietly = TRUE)) {
install.packages("BiocManager")
}
BiocManager::install("spatialLIBD")
```
If you want to use the development version of `spatialLIBD`, you will need to use the R version corresponding to the current Bioconductor-devel branch as described in more detail on the [Bioconductor website](http://bioconductor.org/developers/how-to/useDevel/). Then you can install `spatialLIBD` from GitHub using the following command.
```{r "install_devel", eval = FALSE}
BiocManager::install("LieberInstitute/spatialLIBD")
```
## Access the data
Through the `spatialLIBD` package you can access the processed data in
it’s final R format. However, we also provide a table of links so you
can download the raw data we received from 10x Genomics.
### Processed data
Using `spatialLIBD` you can access the Human DLPFC spatial
transcriptomics data from the 10x Genomics Visium platform. For example,
this is the code you can use to access the layer-level data. For more
details, check the help file for `fetch_data()`.
```{r 'access_data', message=FALSE, fig.height = 8, fig.width = 9, eval = FALSE}
## Load the package
library("spatialLIBD")
## Download the spot-level data
spe <- fetch_data(type = "spe")
## This is a SpatialExperiment object
spe
#> class: SpatialExperiment
#> dim: 33538 47681
#> metadata(0):
#> assays(2): counts logcounts
#> rownames(33538): ENSG00000243485 ENSG00000237613 ... ENSG00000277475
#> ENSG00000268674
#> rowData names(9): source type ... gene_search is_top_hvg
#> colnames(47681): AAACAACGAATAGTTC-1 AAACAAGTATCTCCCA-1 ...
#> TTGTTTCCATACAACT-1 TTGTTTGTGTAAATTC-1
#> colData names(69): sample_id Cluster ... array_row array_col
#> reducedDimNames(6): PCA TSNE_perplexity50 ... TSNE_perplexity80
#> UMAP_neighbors15
#> mainExpName: NULL
#> altExpNames(0):
#> spatialCoords names(2) : pxl_col_in_fullres pxl_row_in_fullres
#> imgData names(4): sample_id image_id data scaleFactor
## Note the memory size
lobstr::obj_size(spe)
#> 2.04 GB
## Remake the logo image with histology information
vis_clus(
spe = spe,
clustervar = "spatialLIBD",
sampleid = "151673",
colors = libd_layer_colors,
... = " DLPFC Human Brain Layers\nMade with research.libd.org/spatialLIBD/"
)
```
<img src="http://research.libd.org/spatialLIBD/reference/figures/README-access_data-1.png" width="600px" align="center" />
### Raw data
You can access all the raw data through [Globus](http://research.libd.org/globus/) (`jhpce#HumanPilot10x`). Furthermore, below you can find the links to the raw data we received from 10x Genomics.
| SampleID | h5_filtered | h5_raw | image_full | image_hi | image_lo | loupe | HTML_report |
|---:|:---|:---|:---|:---|:---|:---|:---|
| 151507 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151507_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151507_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151507_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151507_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151507_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151507.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151507/151507_web_summary.html) |
| 151508 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151508_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151508_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151508_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151508_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151508_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151508.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151508/151508_web_summary.html) |
| 151509 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151509_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151509_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151509_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151509_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151509_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151509.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151509/151509_web_summary.html) |
| 151510 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151510_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151510_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151510_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151510_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151510_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151510.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151510/151510_web_summary.html) |
| 151669 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151669_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151669_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151669_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151669_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151669_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151669.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151669/151669_web_summary.html) |
| 151670 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151670_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151670_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151670_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151670_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151670_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151670.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151670/151670_web_summary.html) |
| 151671 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151671_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151671_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151671_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151671_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151671_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151671.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151671/151671_web_summary.html) |
| 151672 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151672_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151672_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151672_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151672_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151672_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151672.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151672/151672_web_summary.html) |
| 151673 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151673_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151673_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151673_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151673_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151673_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151673.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151673/151673_web_summary.html) |
| 151674 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151674_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151674_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151674_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151674_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151674_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151674.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151674/151674_web_summary.html) |
| 151675 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151675_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151675_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151675_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151675_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151675_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151675.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151675/151675_web_summary.html) |
| 151676 | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151676_filtered_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/h5/151676_raw_feature_bc_matrix.h5) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151676_full_image.tif) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151676_tissue_hires_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/images/151676_tissue_lowres_image.png) | [AWS](https://spatial-dlpfc.s3.us-east-2.amazonaws.com/loupe/151676.cloupe) | [GitHub](https://github.com/LieberInstitute/HumanPilot/blob/master/10X/151676/151676_web_summary.html) |
## Citation
Below is the citation output from using `citation('spatialLIBD')` in R.
Please run this yourself to check for any updates on how to cite
**spatialLIBD**.
```{r 'citation'}
citation('spatialLIBD')
```
# `HumanPilot` code
## Re-shaping your data to our structure
As described in the `spatialLIBD` vignette, you can see the scripts in this repository for re-shaping your data to look like ours. That is.
* `reorganize_folder.R` available [here](https://github.com/LieberInstitute/HumanPilot/blob/master/reorganize_folder.R) re-organizes the raw data we were sent by 10x Genomics.
* `Layer_Notebook.R` available [here](https://github.com/LieberInstitute/HumanPilot/blob/master/Analysis/Layer_Notebook.R) reads in the Visium data and builds a list of `RangeSummarizedExperiment()` objects from `r Biocpkg('SummarizedExperiment')`, one per sample (image) that is eventually saved as `Human_DLPFC_Visium_processedData_rseList.rda`.
* `convert_sce.R` available [here](https://github.com/LieberInstitute/HumanPilot/blob/master/Analysis/convert_sce.R) reads in `Human_DLPFC_Visium_processedData_rseList.rda` and builds an initial `sce` object with image data under `metadata(sce)$image` which is a single data.frame. Subsetting doesn't automatically subset the image, so you have to do it yourself when plotting as is done by `sce_image_clus_p()` and `sce_image_gene_p()`. Having the data from all images in a single object allows you to use the spot-level data from all images to compute clusters and do other similar analyses to the ones you would do with sc/snRNA-seq data. The script creates the `Human_DLPFC_Visium_processedData_sce.Rdata` file.
* `sce_scran.R` available [here](https://github.com/LieberInstitute/HumanPilot/blob/master/Analysis/sce_scran.R) then uses `r Biocpkg('scran')` to read in `Human_DLPFC_Visium_processedData_sce.Rdata`, compute the highly variable genes (stored in our final `sce` object at `rowData(sce)$is_top_hvg`), perform dimensionality reduction (PCA, TSNE, UMAP) and identify clusters using the data from all images. The resulting data is then stored as `Human_DLPFC_Visium_processedData_sce_scran.Rdata` and is the main object used throughout our analysis code `r citep(bib[['spatialLIBDpaper']])`.
* `make-data_spatialLIBD.R` available in the source version of `spatialLIBD` and [online here](https://github.com/LieberInstitute/spatialLIBD/blob/master/inst/scripts/make-data_spatialLIBD.R) is the script that reads in `Human_DLPFC_Visium_processedData_sce_scran.Rdata` as well as some other outputs from our analysis and combines them into the final `sce` and `sce_layer` objects provided by `r Biocpkg('spatialLIBD')` `r citep(bib[['spatialLIBD']])`. This script simplifies some operations in order to simplify the code behind the `r CRANpkg('shiny')` application provided by `r Biocpkg('spatialLIBD')`.
## [10X](10X/) directory
Contains some of the raw files provided by 10X. Given their size, we only included the small ones here.
## [Analysis](Analysis/) directory
The `README.md` was the one we initially prepared for our collaborators at an early stage of the project. That README file described some of our initial explorations using packages such as `r Biocpkg('scran')`, `r Biocpkg('zinbwave')` and other approaches such as using k-means with X/Y spatial information. These analyses were not used for our manuscript beyond creating the `sce` object we previously described.
The 10x Genomics file structure is replicated inside [Histology](Analysis/Histology/) where we saved the image segmentation analysis output to estimate the number of cells per spot. This involved running some [histology image segmentation software](https://www.mathworks.com/help/images/color-based-segmentation-using-k-means-clustering.html) and the [counting code](Analysis/Histology/code) that requires our file structure (`sgeID` input).
The main layer-level analysis code is located at [Layer_Guesses](Analysis/Layer_Guesses), for example [layer_specificity.R](Analysis/Layer_Guesses/layer_specificity.R) is the R script for pseudo-bulking the spot-level data to create the layer-level data. The `spatialLIBD` layer annotation files are saved in the [First_Round](Analysis/Layer_Guesses/First_Round/) and [Second_Round](Analysis/Layer_Guesses/Second_Round) directories which you can upload to the shiny web application.
We also include directories with code for processing external datasets such as [he_layers](Analysis/he_layers), [allen_data](Analysis/allen_data/), [hafner_vglut](Analysis/hafner_vglut/).
We would like to highlight that a lot of the plotting code and functionality from these scripts has been implemented in `r Biocpkg('spatialLIBD')` which would make a lot of our analysis simpler. Finally, for reproducibility purposes we included the R session information in many of our R scripts. Although in general we used R 3.6.1 and 3.6.2 with Bioconductor release 3.10.
## [outputs](outputs/) directory
Contains outputs from the different unsupervised, semi-supervised, known gene marker based and other clustering results. The analysis code that generates these CSV files is located inside R Markdown files at the [Analysis](Analysis/) directory such as [SpatialDE_clustering.Rmd](Analysis/SpatialDE_clustering.Rmd).
<a href="https://www.libd.org/"><img src="http://lcolladotor.github.io/img/LIBD_logo.jpg" width="250px"></a>
# Bibliography
```{r bibliography, results = 'asis', echo = FALSE, warning = FALSE, message = FALSE}
## Print bibliography
bibliography()
```
# Internal
* JHPCE location: `/dcs04/lieber/lcolladotor/with10x_LIBD001/HumanPilot`
* Main `sce` R object file: `/dcs04/lieber/lcolladotor/with10x_LIBD001/HumanPilot/Analysis/Human_DLPFC_Visium_processedData_sce_scran.Rdata`.