T Cell Cytometery Analysis

Immdonor is a GraphQL API of Immport Organ Donor metadata hosted on Hasura Cloud. This API lets you request exactly what data you need from the Immport data model and returns a JSON output which can be converted to a R dataframe!

This is a fast way to link Flow Cytometry Standard (.FCS) files to their metadata and enables more robust analysis with Cytoverse. You will need to have the FCS files saved locally (see quick_fetch.R).

Connect to API

You can connect using your preferred GraphQL R client. Here we use rOpenSci’s ghql package

library(ghql)

#By connecting to this API you agree to the Immport.org Terms of Service
con <- GraphqlClient$new(url = "https://resolved-lab-57.hasura.app/v1/graphql")

GET

qry <- Query$new()

studies<- qry$query('studies','{
  shared_data_study {
    study_accession
    actual_enrollment
    minimum_age
    maximum_age
  }
}')
x <- con$exec(qry$queries$studies)

ANSWER

studies<- as.data.frame (jsonlite::fromJSON(x))

study_accession	actual_enrollment	minimum_age	maximum_age
SDY702	56	3.00	73.00
SDY1041	20	21.00	70.00
SDY1097	43	.20	63.00
SDY1389	51	9.00	76.00

Organ Donor Metadata

The studies listed above feature cytometry data generously donated by organ donors. We can build queries to return exactly what metadata we would like to know about these organ donors.

qry <- Query$new()

files<- qry$query('files','{
  resultFiles {
    filePath
    studyAccession
    subjectAccession
    maxSubjectAge
    gender
    biosampleType
    parameters
    panel_id
    markers
    
  }
}')

x <- con$exec(qry$queries$files)
files<- as.data.frame (jsonlite::fromJSON(x))

Biosample types

Primary Lymphoid Organs	Count
Bone Marrow	150
Thymus	31

Secondary Lymphoid Organs	Count
Inguinal lymph node	102
Lung lymph node	246
Mesenteric lymph node	120
Spleen	182
Tonsil	12

Mucosal	Count
Colon	76
Ileum	61
Jejunum	58
Lung	130
PBMC	136
Whole blood	31

Panels

Here are the top 5 staining panels with the most biosamples

panel_id	parameters	markers	n
FCM-19	14	FITC-A-CD25\|Qdot 605-A-CD45RA\|Qdot 655-A-CD3\|QDot 705-A-CD127\|APC-A-CD19\|PE-Cy7-A-CD4\|PE-A-FOXP3	81
FCM-56	20	APC-A-CMV\|BV570-A-CD3\|Qdot 655-A-CD4\|Qdot 605 UV-A-CD8\|Qdot 605-A-CD45RA\|Alexa Fluor 488-A-CCR7\|QDot 705-A-NA\|BV421-A-CD95\|Alexa Fluor 700-A-INFg\|PerCP-Cy5-5-A-TNFa\|PE-Cy7-A-IL2\|PE-A-CD107a\|DAPI-A-DEAD	78
FCM-21	18	FITC-A-CD25\|PE-Cy5-A-CTLA-4\|BV421-A-CD69\|Qdot 605-A-CD45RA\|Qdot 655-A-CD3\|QDot 705-A-CD127\|APC-A-CD103\|Alexa Fluor 700-A-KI67\|APC-Cy7-A-CD8\|PE-Cy7-A-CD4\|PE-A-FOX-p3	57
FCM-1	14	Alexa Fluor 488-A-CCR7\|PerCP-Cy5-5-A-CD45RO\|BV421-A-CD69\|Qdot 605-A-CD45RA\|Qdot 655-A-CD3\|QDot 705-A-CD127\|APC-A-CD31\|APC-Cy7-A-CD8\|PE-Texas Red-A-CD19\|PE-Cy7-A-CD4\|PE-A-CD28	54
FCM-2	18	Alexa Fluor 488-A-CCR7\|PerCP-Cy5-5-A-CD45RO\|BV421-A-CD69\|Qdot 605-A-CD45RA\|Qdot 655-A-CD3\|QDot 705-A-CD127\|APC-A-CD31\|APC-Cy7-A-CD8\|PE Alexa Fluor 610-A-CD19\|PE-Cy7-A-CD4\|PE-A-CD28	52

T Cell Surface Panel Analysis

panel_id	parameters	markers	n
FCM-2	18	Alexa Fluor 488-A-CCR7\|PerCP-Cy5-5-A-CD45RO\|BV421-A-CD69\|Qdot 605-A-CD45RA\|Qdot 655-A-CD3\|QDot 705-A-CD127\|APC-A-CD31\|APC-Cy7-A-CD8\|PE Alexa Fluor 610-A-CD19\|PE-Cy7-A-CD4\|PE-A-CD28	52
FCM-3	18	Alexa Fluor 488-A-NA\|PerCP-Cy5-5-A-NA\|BV421-A-NA\|Qdot 605-A-NA\|Qdot 655-A-NA\|QDot 705-A-NA\|APC-A-NA\|APC-Cy7-A-NA\|PE Alexa Fluor 610-A-NA\|PE-Cy7-A-NA\|PE-A-NA	27

Link cytoset to metadata

In order to use all of OpenCyto’s features like collapseDataForGating we need to annotate the donor metadata into the cytoset’s pData.

# Load cytoset 
fcs<- files %>% filter(data.resultFiles.panel_id == "FCM-2"|data.resultFiles.panel_id == "FCM-3")
fn<- fcs$data.resultFiles.filePath
cs<- load_cytoset_from_fcs(files = fn)

#Annotate immdonor metadata to pData
p<- pData(cs)
m<- cbind(p,fcs)
pData(cs)<- m

# Harmonize marker names
channels <- colnames(cs)
markers <- as.vector(pData(parameters(cs[[1]]))$desc)
names(markers)<- channels
markernames(cs) <- markers

Compensate and Transform

# Apply file internal compensation
comps <- lapply(cs, function(cf) spillover(cf)$SPILL)
cs_comp <- compensate(cs, comps)

# Transform fluorescent channels with FCSTrans
channels_to_exclude <- c(grep(colnames(cs), pattern="FSC"),
                         grep(colnames(cs), pattern="SSC"),
                         grep(colnames(cs), pattern="Time"))
chnls <- colnames(cs)[-channels_to_exclude]

fcstrans<- FCSTransTransform(transformationId = "defaultFCSTransTransform", channelrange = 2^18, channeldecade = 4.5, range = 4096, cutoff = -111, w = 0.5, rescale = TRUE)
transList <- transformList(chnls, fcstrans)

cs_trans<- transform(cs_comp,transList)

cs<- save_cytoset(cs_trans, "cytosets/tcell)

Build Autogating strategy

Since we annotated the cytoset with each file’s age, gender and biosample type we can use the collapseDataForGating feature and groupby biosampleType to improve our autogating.

cs<- load_cytoset(path = "cytosets/tcell")
gs<- GatingSet(cs)

gs_add_gating_method(gs, alias = "nonDebris",
                     pop = "+",
                     parent = "root",
                     dims = "FSC-A",
                     gating_method = "mindensity",
                     gating_args = "min = 20000, max=50000",
                     collapseDataForGating = "TRUE",
                     groupBy = "data.resultFiles.biosampleType") 
## ...
## Warning in .gating_gtMethod(x, y, ...): NAs introduced by coercion
## done

gs_add_gating_method(gs, alias = "singlets",
                     pop = "+",
                     parent = "nonDebris",
                     dims = "FSC-A,FSC-H",
                     gating_method = "singletGate")
## ...
## done


gs_add_gating_method(gs, alias = "bcells",
                     pop = "+/-",
                     parent = "singlets",
                     dims = "CD19",
                     gating_method = "mindensity",
                     collapseDataForGating = "TRUE",
                     groupBy = "data.resultFiles.biosampleType")
## ...
## Warning in .gating_gtMethod(x, y, ...): NAs introduced by coercion
## done

gs_add_gating_method(gs, alias = "CD4",
                     pop = "+/-+/-",
                     parent = "CD19-",
                     dims = "CD3,CD4",
                     gating_method = "mindensity",
                     gating_args = "min = 1500, max=2500",
                     collapseDataForGating = "TRUE",
                     groupBy = "data.resultFiles.biosampleType")
## ...
## Warning in .gating_gtMethod(x, y, ...): NAs introduced by coercion

## Warning in .gating_gtMethod(x, y, ...): NAs introduced by coercion
## done

gs_add_gating_method(gs, alias = "Tmem",
                     pop = "+/-+/-",
                     parent = "CD4+",
                     dims = "CCR7,CD45RA",
                     gating_method = "mindensity",
                     gating_args = "min = 1750, max=2500",
                     collapseDataForGating = "TRUE",
                     groupBy = "data.resultFiles.biosampleType")
## ...
## Warning in .gating_gtMethod(x, y, ...): NAs introduced by coercion

## Warning in .gating_gtMethod(x, y, ...): NAs introduced by coercion
## done


lln<- subset(gs, data.resultFiles.biosampleType == "Lung lymph node")
lung<- subset(gs, data.resultFiles.biosampleType == "Lung")
spleen<- subset(gs, data.resultFiles.biosampleType == "Spleen")
blood<- subset(gs, data.resultFiles.biosampleType == "Whole blood")

CD3+CD4 T cells

Lung Lymph node

ggcyto(lln, aes(x = "CD4", y = "CD3")) + geom_gate("CD4+") + geom_hex(bins = 128)+ ggcyto_par_set(limits = "instrument")
## Coordinate system already present. Adding new coordinate system, which will replace the existing one.

Lung

ggcyto(lung, aes(x = "CD4", y = "CD3")) + geom_gate("CD4+") + geom_hex(bins = 128)+ ggcyto_par_set(limits = "instrument")
## Coordinate system already present. Adding new coordinate system, which will replace the existing one.

Spleen

ggcyto(spleen, aes(x = "CD4", y = "CD3")) + geom_gate("CD4+") + geom_hex(bins = 128)+ ggcyto_par_set(limits = "instrument")
## Coordinate system already present. Adding new coordinate system, which will replace the existing one.

CD4 memory subsets

Lung Lymph Node

ggcyto(gs_pop_get_data(lln ,"CD4+"), aes(x = "CCR7", y = "CD45RA")) + geom_hex(bins = 128)

Lung

ggcyto(gs_pop_get_data(lung ,"CD4+"), aes(x = "CCR7", y = "CD45RA")) + geom_hex(bins = 128)

Spleen

ggcyto(gs_pop_get_data(spleen ,"CD4+"), aes(x = "CCR7", y = "CD45RA")) + geom_hex(bins = 128)

Metacyto analysis

work in progress…

fcs<- files %>% filter(data.resultFiles.panel_id == "FCM-2"|data.resultFiles.panel_id == "FCM-3")

fcs<- fcs %>%
  dplyr::filter(data.resultFiles.biosampleType == "Lung lymph node"|data.resultFiles.biosampleType == "Lung"|data.resultFiles.biosampleType == "Spleen"|data.resultFiles.biosampleType == "PBMC"|data.resultFiles.biosampleType == "Whole blood"|data.resultFiles.biosampleType == "Inguinal lymph node"|data.resultFiles.biosampleType == "Colon"|data.resultFiles.biosampleType == "Ileum"|data.resultFiles.biosampleType == "Jejunum")

fcs$fcs_files<- fcs$data.resultFiles.filePath
fcs$study_id <- fcs$data.resultFiles.biosampleType

fcs_info<- fcs %>% 
  dplyr::select(fcs_files, study_id)

sample_info <- fcs %>%
  dplyr::select(fcs_files, data.resultFiles.maxSubjectAge,data.resultFiles.biosampleType, data.resultFiles.gender)

preprocessing.batch(inputMeta = fcs_info,
                    assay = "FCM",
                    outpath = "metacyto/panel/preprocess_output",
                    b = 1/150,
                    excludeTransformParameters=c("FSC-A","FSC-W","FSC-H","SSC-A","SSC-W","SSC-H","Time"))
## Study ID =  Colon  Preprocessing 
## Study ID =  Ileum  Preprocessing 
## Study ID =  Jejunum  Preprocessing 
## Study ID =  Whole blood  Preprocessing 
## Study ID =  Lung lymph node  Preprocessing 
## Study ID =  Inguinal lymph node  Preprocessing 
## Study ID =  Lung  Preprocessing 
## Study ID =  Spleen  Preprocessing 
## Study ID =  PBMC  Preprocessing 
## Preprocess result stored in the folder: metacyto/panel/preprocess_output

files=list.files("metacyto/panel",pattern="processed_sample",recursive=TRUE,full.names=TRUE)

cs<- load_cytoset(path = "cytosets/tcell")
new<- markernames(cs)
channels_to_exclude <- c(grep(colnames(cs), pattern="FSC"),
                         grep(colnames(cs), pattern="SSC"),
                         grep(colnames(cs), pattern="Time"))
old <- colnames(cs)[-channels_to_exclude]
old<- toupper(old)
oldplus<-paste0(old,"+")
oldminus<-paste0(old,"-")

nameUpdator(oldNames=old, newNames=new, files=files)
nameUpdator(oldNames=oldplus, newNames=new, files=files)
nameUpdator(oldNames=oldminus, newNames=new, files=files)


#define parameters that we don't want to cluster
excludeClusterParameters=c("FSC-A","FSC-W","FSC-H","SSC-A","SSC-W","SSC-H","Time")

cluster_label=autoCluster.batch(preprocessOutputFolder="metacyto/panel/preprocess_output",
                                excludeClusterParameters=excludeClusterParameters,
                                labelQuantile=0.95,
                                minPercent = 0.05,
                                clusterFunction=flowSOM.MC)
## Clustering , study ID =  Colon
## Building SOM
## Mapping data to SOM
## Building MST
## Clustering , study ID =  Ileum
## Building SOM
## Mapping data to SOM
## Building MST
## Clustering , study ID =  Jejunum
## Building SOM
## Mapping data to SOM
## Building MST
## Clustering , study ID =  Whole blood
## Building SOM
## Mapping data to SOM
## Building MST
## Clustering , study ID =  Lung lymph node
## Building SOM
## Mapping data to SOM
## Building MST
## Clustering , study ID =  Inguinal lymph node
## Building SOM
## Mapping data to SOM
## Building MST
## Clustering , study ID =  Lung
## Building SOM
## Mapping data to SOM
## Building MST
## Clustering , study ID =  Spleen
## Building SOM
## Mapping data to SOM
## Building MST
## Clustering , study ID =  PBMC
## Building SOM
## Mapping data to SOM
## Building MST

searchCluster.batch(preprocessOutputFolder="metacyto/panel/preprocess_output",
                    outpath="metacyto/panel/search_output",
                    clusterLabel=cluster_label)
## Searching , study ID =  Colon
## Searching , study ID =  Ileum
## Searching , study ID =  Jejunum
## Searching , study ID =  Whole blood
## Searching , study ID =  Lung lymph node
## Searching , study ID =  Inguinal lymph node
## Searching , study ID =  Lung
## Searching , study ID =  Spleen
## Searching , study ID =  PBMC

files=list.files("metacyto/panel/search_output",pattern="cluster_stats_in_each_sample",recursive=TRUE,full.names=TRUE)
fcs_stats=collectData(files,longform=TRUE)


# join the cluster summary statistics with sample information
all_data=inner_join(fcs_stats,sample_info,by="fcs_files")

# See the fraction of what clusters are affected by Gender (while controlling for age)
GA=glmAnalysis(value="value",
               variableOfInterst="data.resultFiles.gender",
               parameter="fraction", 
               otherVariables=c("data.resultFiles.maxSubjectAge"),
               studyID="study_id",label="label",
               data=all_data,CILevel=0.95,ifScale=c(TRUE,FALSE))

GA=GA[order(GA$`Effect_size`),]

GA$`label`=as.character(GA$`label`)
w = which((GA$`label`)>15)

## Warning: Use of `GA$label` is discouraged. Use `label` instead.

## Warning: Use of `GA$Effect_size` is discouraged. Use `Effect_size` instead.

Top 20 largest effect size

	label	Effect_size	SE	t_value	p_value	lower	upper	N
47	CD28+\|CD3+\|CD4+\|CD45RO+	0.8031651	0.2278789	3.524526	0.0008629	0.3464856	1.2598446	66
184	CD28+\|CD4+\|CD45RO+	0.7426466	0.2428331	3.058260	0.0034352	0.2559982	1.2292949	66
85	CCR7+\|CD19+\|CD31+\|CD4+\|CD45RO+\|CD8+	0.7358015	0.2545410	2.890700	0.0056343	0.2247892	1.2468138	61
128	CCR7+\|CD19+\|CD3+\|CD31+\|CD4+\|CD45RO+\|CD8+	0.7305022	0.2551306	2.863248	0.0060710	0.2183061	1.2426982	61
205	CD4+\|CD45RO+	0.7237409	0.2444161	2.961102	0.0045181	0.2339201	1.2135616	66
203	CD28+\|CD3+\|CD4+\|CD45RO+\|CD8+	0.7039717	0.2343360	3.004113	0.0041215	0.2335226	1.1744207	61
30	CCR7+\|CD19+	0.6178449	0.2168472	2.849217	0.0061552	0.1832733	1.0524165	66
78	CCR7+\|CD19+\|CD45RO+	0.6075617	0.2304315	2.636626	0.0108627	0.1457666	1.0693569	66
103	CCR7+\|CD19+\|CD31+\|CD4+\|CD45RO+	0.5967158	0.2418360	2.467439	0.0167492	0.1120655	1.0813660	66
11	CD127-\|CD19+\|CD4+\|CD45RO+\|CD8+	0.5930820	0.2192813	2.704663	0.0095541	0.1516916	1.0344725	55
104	CCR7+\|CD19+\|CD31+\|CD4+	0.5903663	0.2410285	2.449363	0.0175242	0.1073343	1.0733983	66
132	CCR7+\|CD3+\|CD45RO+	0.5903451	0.2647291	2.229997	0.0298485	0.0598160	1.1208741	66
74	CCR7+\|CD19+\|CD28+\|CD31+\|CD4+\|CD45RA+\|CD45RO+\|CD8+	0.5826280	0.2760517	2.110576	0.0397328	0.0284311	1.1368248	61
213	CD19+\|CD3+\|CD45RA+\|CD45RO+	0.5573624	0.2245925	2.481660	0.0161614	0.1072689	1.0074559	66
151	CD4+	0.5528965	0.2464158	2.243754	0.0288942	0.0590682	1.0467248	66
17	CCR7+\|CD19+\|CD28+\|CD45RO+\|CD8+	0.5446699	0.2295150	2.373134	0.0214444	0.0838993	1.0054405	61
95	CCR7+\|CD19+\|CD28+\|CD31+\|CD4+\|CD45RA+\|CD8+	0.5394233	0.2465378	2.187994	0.0329366	0.0453505	1.0334961	66
120	CCR7+\|CD19+\|CD28+\|CD3+\|CD4+\|CD45RA+	0.5315369	0.2320631	2.290484	0.0258510	0.0664721	0.9966016	66
136	CD19+\|CD45RA+	0.5297685	0.2182681	2.427146	0.0185209	0.0923495	0.9671875	66
226	CCR7+\|CD19+\|CD28+\|CD3+\|CD31+\|CD4+\|CD45RA+\|CD45RO+\|CD69+\|CD8+	0.5258602	0.2900366	1.813083	0.0763471	-0.0579531	1.1096735	55

Coming soon:
- CyTOF data support
Please contact me with ANY questions, comments or suggestions!!!
- Slack: Join the immunespace slack!
- iMessage: mrjaffery at icloud dot com
- email: mrjaffery at gmail dot com

Name		Name	Last commit message	Last commit date
Latest commit History 43 Commits
R		R
db		db
metacyto/panel		metacyto/panel
readme_files/figure-gfm		readme_files/figure-gfm
.gitignore		.gitignore
DESCRIPTION		DESCRIPTION
Untitled Diagram.drawio		Untitled Diagram.drawio
readme.md		readme.md
readme.rmd		readme.rmd

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

T Cell Cytometery Analysis

Connect to API

GET

ANSWER

Organ Donor Metadata

Biosample types

Panels

T Cell Surface Panel Analysis

Link cytoset to metadata

Compensate and Transform

Build Autogating strategy

CD3+CD4 T cells

CD4 memory subsets

Metacyto analysis

About

Languages

mrj31/immdonor

Folders and files

Latest commit

History

Repository files navigation

T Cell Cytometery Analysis

Connect to API

GET

ANSWER

Organ Donor Metadata

Biosample types

Panels

T Cell Surface Panel Analysis

Link cytoset to metadata

Compensate and Transform

Build Autogating strategy

CD3+CD4 T cells

CD4 memory subsets

Metacyto analysis

About

Topics

Resources

Stars

Watchers

Forks

Languages