cytoqc – qc for keywords

library(flowCore)
library(flowWorkspace)
# library(cytoqc)
devtools::load_all("../")

Load the FCS

files <- list.files(data_dir, ".fcs", full.names = TRUE)
cqc_data <- cqc_load_fcs(files)
cqc_data

## cytoqc data: 
## 35  samples

Group by keywords

groups <- cqc_group(cqc_data, "keyword")
su <- summary(groups)
knit_print(su)

group_id	nFCS	keyword
6	14	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, FILENAME, GUID, Instr. Sett. File, Mixing Vol, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, Patient ID, Sample ID, SAMPLE ID, Sample Vol, TIMETICKS, transformation
2	11	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, END BATCH, FILENAME, GUID, Instr. Sett. File, Mixing Vol, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, PANEL NAME, Patient ID, Sample ID, SAMPLE ID, Sample Vol, START BATCH, TIMETICKS, transformation
4	7	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, END BATCH, FILENAME, GUID, Instr. Sett. File, Mixing Vol, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, PANEL NAME, pt name, Sample ID, SAMPLE ID, Sample Vol, START BATCH, TIMETICKS, transformation
1	1	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, END BATCH, FILENAME, GUID, Instr. Sett. File, Mixing Vol, new_kw, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, PANEL NAME, patient id, Sample ID, SAMPLE ID, Sample Vol, START BATCH, TIMETICKS, transformation
3	1	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, END BATCH, FILENAME, GUID, Instr. Sett. File, Mixing Vol, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, PANEL NAME, PatientId, Sample ID, SAMPLE ID, Sample Vol, START BATCH, TIMETICKS, transformation
5	1	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, END BATCH, FILENAME, GUID, Instr. Sett. File, Mixing Vol, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, PANEL NAME, Sample ID, SAMPLE ID, Sample Vol, START BATCH, TIMETICKS, transformation

Match against reference and report the discrepancy

match_result <- cqc_match_reference(groups, ref = 2, select = c(1,4,3,5))
match_result

group_id	Not in reference	Missing channels
1	new_kw,patient id	Patient ID
3	PatientId	Patient ID
4	pt name	Patient ID
5		Patient ID

Proposed change
patient id –> Patient ID
new_kw
PatientId –> Patient ID

Apply the fix

cqc_fix(solution)

Drop the group that has missing keyword

groups <- cqc_drop_groups(groups, 5)

Inspect the content of the two keywords that are likely to be semantically the same

sub_data_ref <- cqc_get_data(groups, id = 2)
unique(sapply(sub_data_ref, keyword, keyword = "Patient ID"))

## [[1]]
## [1] "#10 MS 240816"
## 
## [[2]]
## [1] "#9 FP 242399"

sub_data <- cqc_get_data(groups, id = 4)
unique(sapply(sub_data, keyword, keyword = "pt name"))

## [[1]]
## [1] "BJ#230500"

Consolidate the two keywords if they are considered to be the same

invisible(lapply(sub_data, cf_keyword_rename, "pt name", "Patient ID"))

Refresh QC report

cqc_data <- cqc_get_data(groups)
groups <- cqc_group(cqc_data, "keyword")
knit_print(summary(groups))

group_id	nFCS	keyword
1	20	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, END BATCH, FILENAME, GUID, Instr. Sett. File, Mixing Vol, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, PANEL NAME, Patient ID, Sample ID, SAMPLE ID, Sample Vol, START BATCH, TIMETICKS, transformation
2	14	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, FILENAME, GUID, Instr. Sett. File, Mixing Vol, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, Patient ID, Sample ID, SAMPLE ID, Sample Vol, TIMETICKS, transformation

Check the difference

res <- cqc_match_reference(groups, ref = 2, select = 1)
res

group_id	Not in reference	Missing channels
1	END BATCH,PANEL NAME,START BATCH

split into two groups if they can’t be merged

grps <- split(groups)
grps

## $`1`
## cytoqc data: 
## 20  samples 
## 
## $`2`
## cytoqc data: 
## 14  samples

or drop these keywords if they are not critical

solution <- cqc_find_solution(res)
solution

Proposed change
END BATCH
PANEL NAME
START BATCH

cqc_fix(solution)
knit_print(summary(cqc_group(cqc_data, "keyword")))

group_id	nFCS	keyword
1	34	Acquisition Doc., Analysis Doc., CalibFile, CASE NUMBER, CYTNUM, Data File Prefix Part #1, Data File Prefix Part #2, Data File Prefix Part #3, Day, FILENAME, GUID, Instr. Sett. File, Mixing Vol, Number of Mixes, Number of Washes, NUMBER WELL INFO KEYWORDS, Patient ID, Sample ID, SAMPLE ID, Sample Vol, TIMETICKS, transformation

A QC tool for openCyto

cytoqc – qc for keywords

Load the FCS

Group by keywords

Match against reference and report the discrepancy

Recommend the fix for the selected groups

Apply the fix

Drop the group that has missing keyword

Inspect the content of the two keywords that are likely to be semantically the same

Consolidate the two keywords if they are considered to be the same

Refresh QC report

Check the difference

split into two groups if they can’t be merged

or drop these keywords if they are not critical