There are different data formats and structures that can be exported from the AtoMx™ Spatial Informatics Portal (SIP). These include the raw data (i.e., with CellStatsDir, RunSummary, and AnalysisResults folders), Seurat (with or without images), Tiledb, and “flat files”. The flat files get their name because they are in a human-readable and accessible format (i.e., comma separated files). These files, like Seurat files and Tiledb files, aren’t actually raw data but are processed data and can include additional analysis results (e.g., cell typing data).
The first use of the flat file format was about a year and a half ago when He et al. (2022) released the first public CosMx™ SMI dataset, consisting of ~800k cells with 980 RNA targets from multiple tissues of NSCLC FFPE. Now there are additional public datasets that span two species (mouse, human), four tissues (lung, liver, brain, pancreas), and three levels of plex (1k, 6k, whole transcriptome). Because the flat files generated in AtoMx SIP differ from the flat files in these public data releases, I thought it might be helpful to show a comparison.
Specifically, in this post I compare the flat files created for Lung5_rep1 of the NSCLC dataset side-by-side with the updated (AtoMx 1.3.2) flat file format from an unrelated tissue. The specific values will be different since they are different datasets, of course, but the following side-by-sides show similarities and differences between the formats.
1 FOV positions flat file
When we compare the FOV positions between old and new, you can see that the column name ‘fov’ has been changed to ‘FOV’ and the newer format includes global positions of the FOVs in units of mm in addition to pixels.
| Column Name | Type | Description |
|---|---|---|
| fov\(^{Δ}\) | Int | The field of view (FOV) number |
| x_global_px | float | The x location (in pixels) of the FOV relative to other FOVs |
| y_global_px | float | The y location (in pixels) of the FOV relative to other FOVs |
Example:
fov x_global_px y_global_px
1 3188.889 155216.7
2 8661.111 155216.7
3 14133.333 155216.7
4 19605.556 155216.7
5 25077.778 155216.7
6 3188.889 158866.7
7 8661.111 158866.7
8 14133.333 158866.7
9 19605.556 158866.7
10 25077.778 158866.7| Column Name | Type | Description |
|---|---|---|
| FOV\(^{Δ}\) | Int | The field of view (FOV) number |
| x_global_px | float | The x location (in pixels) of the FOV relative to other FOVs |
| y_global_px | float | The y location (in pixels) of the FOV relative to other FOVs |
| x_global_mm\(^{*}\) | float | The x location (in millimeters) of the FOV relative to other FOVs |
| y_global_mm\(^{*}\) | float | The y location (in millimeters) of the FOV relative to other FOVs |
FOV x_global_px y_global_px x_global_mm y_global_mm
1 0 29791 0.0000000 3.583410
2 4255 29791 0.5119157 3.583410
3 8511 29791 1.0238314 3.583410
4 12767 29791 1.5357471 3.583410
5 17023 29791 2.0476628 3.583410
6 21279 29791 2.5595785 3.583410
7 25535 29791 3.0714942 3.583410
8 29791 29791 3.5834099 3.583410
9 0 25535 0.0000000 3.071494
10 4255 25535 0.5119157 3.0714942 Expression Matrix
For expression matrices, we see that NegPrb(\d+) is changed to Negative(\d+) and the newer format has SystemControls.
| Column Name | Type | Description |
|---|---|---|
| fov | Int | The field of view (FOV) number |
| cell_ID | Int | Unique identifier for a single cell within a given FOV. To make a unique identifier for a cell within the whole sample use both the fov and cell_ID columns. All transcripts not assigned to a cell are show with a cell_ID value of 0. |
| (Gene expression target) | Int | The number of transcripts observed for a given gene target for a given cell. |
| (Negative Probe; e.g. NegPrb1)\(^{Δ}\) | Int | Negative probes, which do not match any sequence within the transcriptome or genome. These can be used to assess background levels. |
Example:
fov cell_ID AATK ... ZFP36 NegPrb3 ... NegPrb23
1 0 15 ... 184 12 ... 13
1 1 0 ... 0 0 ... 0
1 2 0 ... 1 0 ... 0
1 3 0 ... 0 0 ... 0
1 4 0 ... 0 0 ... 0
1 5 0 ... 0 0 ... 0
1 6 0 ... 0 0 ... 0
1 7 0 ... 1 0 ... 0
1 8 0 ... 0 0 ... 0
1 9 0 ... 0 0 ... 0| Column Name | Type | Description |
|---|---|---|
| fov | Int | The field of view (FOV) number |
| cell_ID | Int | Unique identifier for a single cell within a given FOV. To make a unique identifier for a cell within the whole sample use both the fov and cell_ID columns. All transcripts not assigned to a cell are show with a cell_ID value of 0. |
| (Gene expression target) | Int | The number of transcripts observed for a given gene target for a given cell. |
| (Negative Probe, e.g., Negative1)\(^{Δ}\) | Int | Negative probes, which do not match any sequence within the transcriptome or genome. These can be used to assess background levels. |
| (System Control)\(^{*}\) | Int | System Control codes are codes which do not have any physical probe associated with them. |
Example:
fov cell_ID A1BG ... ZZZ3 Negative1 ... Negative9 SystemControl1 ... SystemControl99
1 1 0 ... 0 0 ... 0 0 ... 0
1 2 0 ... 0 0 ... 0 0 ... 0
1 3 0 ... 0 0 ... 0 0 ... 0
1 4 1 ... 0 0 ... 0 0 ... 0
1 5 0 ... 0 0 ... 0 0 ... 0
1 6 0 ... 0 0 ... 0 0 ... 0
1 7 0 ... 0 0 ... 0 0 ... 0
1 8 0 ... 0 0 ... 0 0 ... 0
1 9 1 ... 0 0 ... 0 0 ... 0
1 10 0 ... 0 0 ... 0 0 ... 03 Metadata file
There are several differences to the metadata files between the legacy and current versions and I’ll highlight a few new additions below. One thing to note is that exported data from AtoMx SIP can have columns with analysis results in addition to the “baseline” columns.
Cell shape metrics – In the legacy version, basic cell shape was described with Area, Width, Height, and AspectRaio (Table 5). The new version of the metadata includes these plus four additional metrics that describe the cell shape (Table 6). These are perimeter, circularity, eccentricity, and solidity. Perimeter is simply the perimeter of the cell in pixels and the latter three are defined in Fu et al. (2024).
SplitRatioToLocal – for cells that are adjacent to the FOV boundaries, the SplitRatioToLocal metric measures the cell area relative to the mean area of cells in the FOVs. For 0 < SplitRatioToLocal < 1, the cell is smaller than average and for SplitRatioToLocal > 1 the cell is larger than average. Note that a value of 0 means the cell is not along the border.
FOV-level metrics – there are some columns that are added as FOV-level metrics. For example, median_RNA provides the median RNA target probe expression across all cells within a given FOV.
| Column Name | Type | Description |
|---|---|---|
| fov | Int | The field of view (FOV) number. |
| cell_ID | Int | Unique identifier for a single cell within a given FOV. To make a unique identifier for a cell within the whole sample use both the fov and cell_ID columns. |
| Area | Int | Number of pixels assigned to a given cell. |
| AspectRatio | float | Width divided by height. |
| CenterX_local_px | Int | The x position of this cell within the FOV, measured in pixels. The pixel edge length is 120 nm. Thus, to convert to microns multiply the pixel value by 0.12028 \(\mu\)m per pixel. |
| CenterY_local_px | Int | Same as CenterX_local_px but for the y dimension. |
| CenterX_global_px | float | See CenterX_local_px description. The global positions describes the relative position of this cell within the large sample reference frame. |
| CenterY_global_px | float | Same as CenterX_global_px but for the y dimension. |
| Width | Int | Cell’s maximum length in x dimension (pixels). |
| Height | Int | Cell’s maximum length in y dimension (pixels). |
| Mean.(IF) | Int | The mean fluorescence intensity for a given cell. |
| Max.(IF) | Int | The max fluorescence intensity for a given cell. |
Example:
'data.frame': 6 obs. of 20 variables:
$ fov : int 1 1 1 1 1 1
$ cell_ID : int 1 2 3 4 5 6
$ Area : int 1259 3723 2010 3358 1213 2647
$ AspectRatio : num 1.34 1.45 1.62 0.47 1 1.38
$ CenterX_local_px : int 1027 2904 4026 4230 4258 66
$ CenterY_local_px : int 3631 3618 3627 3597 3629 3622
$ CenterX_global_px : num 4216 6093 7215 7419 7447 ...
$ CenterY_global_px : num 158848 158835 158844 158814 158846 ...
$ Width : int 47 87 68 48 38 72
$ Height : int 35 60 42 102 38 52
$ Mean.MembraneStain: int 3473 3895 2892 6189 8138 5713
$ Max.MembraneStain : int 7354 13832 6048 16091 19281 12617
$ Mean.PanCK : int 715 18374 3265 485 549 1220
$ Max.PanCK : int 5755 53158 37522 964 874 5107
$ Mean.CD45 : int 361 260 378 679 566 433
$ Max.CD45 : int 845 1232 908 2322 1242 957
$ Mean.CD3 : int 22 13 19 5 17 11
$ Max.CD3 : int 731 686 654 582 674 547
$ Mean.DAPI : int 4979 1110 10482 6065 3311 4151
$ Max.DAPI : int 26374 13229 33824 39512 30136 19269
Note
If analysis has been performed on AtoMx SIP prior to export, additional columns not presented here may be added to the metadata file. For example, if cell typing has been performed, there may be a column named RNA_nbclust_[GUID]_1_clusters containing the estimated cell type.
| Column Name | Type | Description |
|---|---|---|
| fov | Int | The field of view (FOV) number. |
| Area | Int | Number of pixels assigned to a given cell. |
| AspectRatio | float | Width divided by height. |
| CenterX_local_px | Int | The x position of this cell within the FOV, measured in pixels. The pixel edge length is 120 nm. Thus, to convert to microns multiply the pixel value by 0.12028 \(\mu\)m per pixel. |
| CenterY_local_px | Int | Same as CenterX_local_px but for the y dimension. |
| Width | Int | Cell’s maximum length in x dimension (pixels). |
| Height | Int | Cell’s maximum length in y dimension (pixels). |
| Mean.(IF) | Int | The mean fluorescence intensity for a given cell. |
| Max.(IF) | Int | The max fluorescence intensity for a given cell. |
| SplitRatioToLocal\(^{*}\) | float | If cell abuts the FOV border: the ratio of Area to mean cell area for that FOV. If cell does not border the FOV boundary: 0. |
| NucArea\(^{*}\) | Int | Number of pixels assigned to a given nucleus. |
| NucAspectRatio\(^{*}\) | float | Width divided by height of nucleus. |
| Circularity\(^{*}\) | float | Area to perimeter ratio. 1 = circle; < 1 less circular (Fu et al. 2024). |
| Eccentricity\(^{*}\) | float | A cell’s minor axis divided by its major axis (Fu et al. 2024). |
| Perimeter\(^{*}\) | Int | The perimeter of the cell (in pixels) |
| Solidity\(^{*}\) | float | The Area of the cell divided by its convex area. A measure of the “density” of a cell with values < 1 indicating increased cell irregularity (Fu et al. 2024) |
| cell_id\(^{*}\) | string | A study-wide unique cell identifier. Combination of c(ell), slide_ID, fov, and cell_ID. Note that this is equivalent to cell_ID in napari-cosmx. |
| assay_type\(^{*}\) | string | The assay type (Protein or RNA) |
| version\(^{*}\) | string | The version of the target decoding used. |
| Run_Tissue_name\(^{*}\) | string | The name of the slide. |
| Panel\(^{*}\) | string | The panel that was assayed. |
| cellSegmentationSetId\(^{*}\). | string | The cell segmentation set ID. |
| cellSegmentationSetName\(^{*}\) | string | The cell segmentation set name. |
| slide_ID\(^{*}\) | Int | Unique identifier for the slide. |
| CenterX_global_px | float | See CenterX_local_px description. The global positions describes the relative position of this cell within the large sample reference frame. |
| CenterY_global_px | float | Same as CenterX_global_px but for the y dimension. |
| cell_ID | Int | Unique identifier for a single cell within a given FOV. To make a unique identifier for a cell within the whole sample use both the fov and cell_ID columns. |
| unassignedTranscripts\(^{*}\) | float | Proportion of transcripts in the FOV the cell resides in that are not assigned within any cell. This value is an FOV-level metric that is repeated for each cell (excluding cell 0). |
| median_RNA\(^{*}\) | float | FOV-level statistic. Median RNA target probe expression across all cells within a given FOV. |
| RNA_quantile_(proportion)\(^{*}\) | float | FOV-level statistic. The (proportion*100) percentile of RNA target expression across all cells within a given FOV. |
| nCount_RNA\(^{*}\) | Int | Total RNA transcripts observed. |
| nFeature_RNA\(^{*}\) | Int | Total number of unique RNA transcripts observed. |
| median_negprobes\(^{*}\) | float | FOV-level statistic. Median negative probe expression across all cells within a given FOV. |
| negprobes_quantile_(proportion)\(^{*}\) | float | FOV-level statistic. The (proportion*100) percentile of negative probe expression across all cells within a given FOV. |
| nCount_negprobes\(^{*}\) | Int | Total Negative Control Probe counts observed. |
| nFeature_negprobes\(^{*}\) | Int | Total number of unique Negative Control Probe counts observed. |
| median_falsecode\(^{*}\) | float | FOV-level statistic. Median System Control counts across all cells within a given FOV. |
| falsecode_quantile_(proportion)\(^{*}\) | float | FOV-level statistic. The (proportion*100) percentile of System Control counts across all cells within a given FOV. |
| nCount_falsecode\(^{*}\) | Int | Total System Control codes counts observed. These codes do not have a physical probe in the experiment. |
| nFeature_falsecode\(^{*}\) | Int | Total number of unique System Control codes counts observed. |
| Area.um2\(^{*}\) | float | The cell area in units of \(\mu m^{2}\) |
| cell\(^{*}\) | string | Redundant with cell_id |
Example:
'data.frame': 6 obs. of 65 variables:
$ fov : int 1 1 1 1 1 1
$ Area : int 3037 8790 5552 5822 4008 3603
$ AspectRatio : num 0.67 0.95 0.77 0.9 0.97 0.88
$ CenterX_local_px : int 3938 2741 3888 4214 4137 4163
$ CenterY_local_px : int 25 52 57 73 152 187
$ Width : int 76 110 96 81 78 83
$ Height : int 51 104 74 90 76 73
$ Mean.B : int 41 425 88 197 73 91
$ Max.B : int 252 2308 952 604 552 364
$ Mean.G : int 50 1270 154 48 22 26
$ Max.G : int 1192 6960 4192 184 380 160
$ Mean.Y : int 106 366 254 235 97 196
$ Max.Y : int 1228 2864 3884 1340 828 904
$ Mean.R : int 36 62 221 20 4 16
$ Max.R : int 2360 1044 5604 104 60 232
$ Mean.DAPI : int 92 37 181 288 237 324
$ Max.DAPI : int 408 236 924 1060 800 928
$ SplitRatioToLocal : num 0.7 2.01 0 1.33 0 0
$ NucArea : int 1252 0 1180 2384 1284 1536
$ NucAspectRatio : num 0.77 0 1 0.94 0.95 0.81
$ Circularity : num 0.92 1.05 1.03 1.03 0.94 0.83
$ Eccentricity : num 0.76 0.82 0.79 0.71 0.9 0.82
$ Perimeter : int 204 324 260 266 231 233
$ Solidity : num 14.9 27.1 21.4 21.9 17.4 ...
$ cell_id : chr "c_1_1_1" "c_1_1_2" "c_1_1_3" "c_1_1_4" ...
$ assay_type : chr "RNA" "RNA" "RNA" "RNA" ...
$ version : chr "v6" "v6" "v6" "v6" ...
$ Run_Tissue_name : chr "example_tissue" "example_tissue" "example_tissue" "example_tissue" ...
$ Panel : chr "Human RNA 6k Discovery" "Human RNA 6k Discovery" "Human RNA 6k Discovery" "Human RNA 6k Discovery" ...
$ cellSegmentationSetId : chr " a343598a-ed40-4a93-a655-49bc7688021d" " a343598a-ed40-4a93-a655-49bc7688021d" " a343598a-ed40-4a93-a655-49bc7688021d" " a343598a-ed40-4a93-a655-49bc7688021d" ...
$ cellSegmentationSetName: chr " Initial Segmentation" " Initial Segmentation" " Initial Segmentation" " Initial Segmentation" ...
$ slide_ID : int 1 1 1 1 1 1
$ CenterX_global_px : int 21057 19860 21007 21333 21256 21282
$ CenterY_global_px : int 68070 68043 68038 68022 67943 67908
$ cell_ID : int 1 2 3 4 5 6
$ unassignedTranscripts : num 0.0349 0.0349 0.0349 0.0349 0.0349 ...
$ median_RNA : int 86 86 86 86 86 86
$ RNA_quantile_0.75 : int 126 126 126 126 126 126
$ RNA_quantile_0.8 : int 139 139 139 139 139 139
$ RNA_quantile_0.85 : int 157 157 157 157 157 157
$ RNA_quantile_0.9 : int 182 182 182 182 182 182
$ RNA_quantile_0.95 : int 240 240 240 240 240 240
$ RNA_quantile_0.99 : num 512 512 512 512 512 ...
$ nCount_RNA : int 138 295 234 344 230 249
$ nFeature_RNA : int 86 182 152 217 148 132
$ median_negprobes : int 9 9 9 9 9 9
$ negprobes_quantile_0.75: int 126 126 126 126 126 126
$ negprobes_quantile_0.8 : int 139 139 139 139 139 139
$ negprobes_quantile_0.85: int 157 157 157 157 157 157
$ negprobes_quantile_0.9 : int 182 182 182 182 182 182
$ negprobes_quantile_0.95: int 240 240 240 240 240 240
$ negprobes_quantile_0.99: num 512 512 512 512 512 ...
$ nCount_negprobes : int 0 0 1 0 0 0
$ nFeature_negprobes : int 0 0 1 0 0 0
$ median_falsecode : int 4 4 4 4 4 4
$ falsecode_quantile_0.75: int 126 126 126 126 126 126
$ falsecode_quantile_0.8 : int 139 139 139 139 139 139
$ falsecode_quantile_0.85: int 157 157 157 157 157 157
$ falsecode_quantile_0.9 : int 182 182 182 182 182 182
$ falsecode_quantile_0.95: int 240 240 240 240 240 240
$ falsecode_quantile_0.99: num 512 512 512 512 512 ...
$ nCount_falsecode : int 1 0 0 1 1 1
$ nFeature_falsecode : int 1 0 0 1 1 1
$ Area.um2 : num 43.9 127.2 80.3 84.2 58 ...
$ cell : chr "c_1_1_1" "c_1_1_2" "c_1_1_3" "c_1_1_4" ...4 Transcript coordinates file
Main differences between versions:
- The contents of the
CellCompcolumn differ between version. In the current version “None” replaces “0”. The other three regions–Membrane, Nuclear, Cytoplasm–are unchanged. cellcolumn is added to the newer version.
| Column Name | Type | Description |
|---|---|---|
| fov | Int | The field of view (FOV) number. |
| cell_ID | Int | Unique identifier for a single cell within a given FOV. To make a unique identifier for a cell within the whole sample use both the fov and cell_ID columns. |
| x_global_px | float | The x position (in pixels) relative to the tissue. |
| y_global_px | float | The y position (in pixels) relative to the tissue. |
| x_local_px | float | The x position (in pixels) relative to the given FOV. |
| y_local_px | float | The y position (in pixels) relative to the given FOV. |
| z | Int | The z plane. |
| target | string | The name of the target. |
| CellComp‡ | string | Subcellular location of target. |
Example:
fov cell_ID x_global_px y_global_px x_local_px y_local_px z target CellComp
1 0 6757.402 158836.4 3568.513 3619.7375 11 NEAT1 0
1 0 5111.389 156060.2 1922.500 843.5334 11 NEAT1 0
1 0 7860.461 157809.3 4671.572 2592.6715 11 CCR2 0
1 0 3790.489 155553.9 601.600 337.2168 11 HLA-DRA 0
1 0 3290.639 158023.6 101.750 2806.9750 11 HLA-DRA Membrane
1 0 7020.160 158656.3 3831.271 3439.6000 11 VHL 0
1 0 4252.914 157003.0 1064.025 1786.3376 11 FZD5 Nuclear
1 0 5987.309 157572.5 2798.420 2355.8000 11 CD37 0
1 0 5586.849 157774.2 2397.960 2557.5599 11 ATG12 Membrane| Column Name | Type | Description |
|---|---|---|
| fov | Int | The field of view (FOV) number. |
| cell_ID | Int | Unique identifier for a single cell within a given FOV. To make a unique identifier for a cell within the whole sample use both the fov and cell_ID columns. |
| cell\(^{*}\) | string | A study-wide unique cell identifier. Combination of c(ell), slide ID, fov, and cell_ID. Note that this is equivalent to cell_ID in napari-cosmx. |
| x_local_px | float | The x position (in pixels) relative to the given FOV. |
| y_local_px | float | The y position (in pixels) relative to the given FOV. |
| x_global_px | float | The x position (in pixels) relative to the tissue. |
| y_global_px | float | The y position (in pixels) relative to the tissue. |
| z | Int | The z plane. |
| target | string | The name of the target. |
| CellComp‡ | string | Subcellular location of target. |
Example:
fov cell_ID cell x_local_px y_local_px x_global_px y_global_px z target CellComp
30 4755 c_1_30_4755 29337.6332465278 173483.2 4259.8555 16.57764 3 B2M Cytoplasm
30 4755 c_1_30_4755 29340.4174262153 173488.4 4262.6396 21.71997 3 COL3A1 Cytoplasm
30 4757 c_1_30_4757 29593.4975043403 173480.9 4515.7197 14.27002 8 RPL32 Cytoplasm
30 4759 c_1_30_4759 25211.3444434272 173477.6 133.5667 10.95020 6 COL1A1 Cytoplasm
30 4759 c_1_30_4759 25224.5611029731 173483.9 146.7833 17.28320 6 COL1A2 Cytoplasm
30 4760 c_1_30_4760 25902.6278143989 173480.8 824.8500 14.11694 7 TPSAB1 Cytoplasm
30 4760 c_1_30_4760 25924.0527411567 173477.8 846.2750 11.10010 1 HSP90AB1 Cytoplasm
30 4760 c_1_30_4760 25925.7694159614 173478.0 847.9916 11.34155 6 GLUL Cytoplasm
30 4760 c_1_30_4760 25914.4152899848 173478.9 836.6375 12.27515 6 ADGRE2 Cytoplasm
30 4760 c_1_30_4760 25902.9277411567 173480.7 825.1500 14.07520 8 TPSAB1 Cytoplasm 5 Polygons file
The polygons file was added to the list of flat files and shows the vertices of each cell’s polygon.
(Not applicable)
| Column Name | Type | Description |
|---|---|---|
| fov | Int | The field of view (FOV) number. |
| cell_ID | Int | Unique identifier for a single cell within a given FOV. To make a unique identifier for a cell within the whole sample use both the fov and cell_ID columns. |
| cell | string | A study-wide unique cell identifier. Combination of c(ell), slide ID, and cell_ID. Note that this is equivalent to cell_ID in napari-cosmx. |
| x_local_px | float | The x position (in pixels) of vertex relative to the given FOV. |
| y_local_px | float | The y position (in pixels) of vertex relative to the given FOV. |
| x_global_px | float | The x position (in pixels) of vertex relative to the tissue. |
| y_global_px | float | The y position (in pixels) of vertex relative to the tissue. |
Example:
This example below shows the vertices of cell c_1_2_3.
fov cellID cell x_local_px y_local_px x_global_px y_global_px
2 3 c_1_2_3 279 0 4535 29792
2 3 c_1_2_3 279 1 4535 29791
2 3 c_1_2_3 270 15 4526 29777
2 3 c_1_2_3 266 20 4522 29772
2 3 c_1_2_3 234 53 4490 29739
2 3 c_1_2_3 223 64 4479 29728
2 3 c_1_2_3 214 71 4470 29721
2 3 c_1_2_3 210 72 4466 29720
2 3 c_1_2_3 199 72 4455 29720
2 3 c_1_2_3 186 66 4442 29726
2 3 c_1_2_3 182 64 4438 29728
2 3 c_1_2_3 179 62 4435 29730
2 3 c_1_2_3 176 31 4432 29761
2 3 c_1_2_3 176 4 4432 29788
2 3 c_1_2_3 179 0 4435 29792References
Fu, Xiaohang, Yingxin Lin, David M. Lin, Daniel Mechtersheimer, Chuhan Wang, Farhan Ameen, Shila Ghazanfar, Ellis Patrick, Jinman Kim, and Jean Y. H. Yang. 2024. “BIDCell: Biologically-Informed Self-Supervised Learning for Segmentation of Subcellular Spatial Transcriptomics Data.” Nature Communications 15 (January): 509. https://doi.org/10.1038/s41467-023-44560-w.
He, Shanshan, Ruchir Bhatt, Carl Brown, Emily A Brown, Derek L Buhr, Kan Chantranuvatana, Patrick Danaher, et al. 2022. “High-Plex Imaging of RNA and Proteins at Subcellular Resolution in Fixed Tissue by Spatial Molecular Imaging.” Nature Biotechnology 40 (December): 1794–1806. https://doi.org/10.1038/s41587-022-01483-z.