From f99bf3ec883079ec5f5bd8b6540d8c17b3ed1068 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Todor=20Kondi=C4=87?= <kontrapunkt@uclmail.net>
Date: Tue, 28 Mar 2023 15:49:37 +0200
Subject: [PATCH] extract_spectra: Extract full MS2 spectra.
---
R/api.R | 18 ++++++++----------
R/data-model.R | 10 ++++++++++
R/extraction.R | 29 +++++++++++++++++++++++++++++
3 files changed, 47 insertions(+), 10 deletions(-)
diff --git a/R/api.R b/R/api.R
index a0555e6..44d1ab1 100644
--- a/R/api.R
+++ b/R/api.R
@@ -374,24 +374,22 @@ extr_data <-function(m) {
idx=integer(0),
rt=numeric(0),
intensity=numeric(0))
+
+ spectra = empty_spectra_table()
+
for (fn in names(lfdata)) {
- ## x = lfdata[[fn]]$ms2[cgram_ms1,
- ## .(an,
- ## ce,
- ## precid=i.precid,
- ## rt,
- ## intensity),
- ## on=c(prec_idx="idx"),
- ## nomatch=NULL,
- ## allow.cartesian=T]
rtab = relate_ms2_to_precid(coarse=coarse,ms2=lfdata[[fn]]$ms2,cgram_ms1=cgram_ms1)
+ sptab = extract_spectra(lms[[fn]],rtab)
cgram_ms2 = rbind(cgram_ms2,rtab)
+ spectra = rbind(spectra,sptab)
}
setkey(cgram_ms1,precid,rt)
setkey(cgram_ms2,precid,ce,rt)
+ setkey(spectra,precid,scan)
m$db$extr$cgm$ms1 = cgram_ms1
m$db$extr$cgm$ms2 = cgram_ms2
- browser()
+ m$db$extr$spectra = spectra
+
m
}
diff --git a/R/data-model.R b/R/data-model.R
index 4d9876e..79f177d 100644
--- a/R/data-model.R
+++ b/R/data-model.R
@@ -103,3 +103,13 @@ make_db_precursors <- function(m) {
m$db$precursors = masses
m
}
+
+
+empty_spectra_table <- function() {
+ r = data.table(precid=integer(0),
+ scan=character(0),
+ mz=numeric(0),
+ intensity=numeric(0))
+ setkey(r,precid,scan)
+ r
+}
diff --git a/R/extraction.R b/R/extraction.R
index aa5c516..c49e959 100644
--- a/R/extraction.R
+++ b/R/extraction.R
@@ -700,8 +700,19 @@ get_fdata <- function(ms) {
relate_ms2_to_precid <- function(coarse,ms2,cgram_ms1) {
+ ## Take `coarse' table (the one with coarse mass limits), ms2
+ ## fData and ms1 chromatogram, then relate precids from cgram_ms1
+ ## to ms2 data.
+
+ ## Select those MS2 entries the parents of which coarsely match
+ ## compound lists masses.
res = ms2[coarse,on=.(prec_mz>iso_coarse_min,prec_mz<iso_coarse_max),.(prec_mz=x.prec_mz,precid,prec_idx,scan,idx,ce,rt,intensity),nomatch=NULL]
setkey(res,precid,prec_idx)
+
+ ## Now, make sure those coarsely matched MS2 actually have a
+ ## parent that finely matches something in the chromatogram (and
+ ## this is by ensuring that a `precid' with the correct scan (idx)
+ ## shows up in the chromatogram.
cgram_ms1[res,on=.(precid,idx==prec_idx),
.(precid,ce,scan=i.scan,
idx=i.idx,rt=i.rt,
@@ -709,3 +720,21 @@ relate_ms2_to_precid <- function(coarse,ms2,cgram_ms1) {
}
+extract_spectra <- function(ms,cgram_ms2) {
+ ## This will extract full MS2 spectra based on ms2 chromatogram entries.
+ indices = cgram_ms2[,.SD,.SDcol=c("precid","scan","idx")]
+
+ res = empty_spectra_table()
+ selind = indices[,unique(.SD),.SDcol=c("scan","idx")]
+ sel = ms[selind$idx]
+
+ masses = mz(sel)
+ intensities = intensity(sel)
+ res = selind
+ setkey(res,scan)
+ res = res[,data.table(mz=masses[[scan]],
+ intensity=intensities[[scan]]),
+ keyby=c("scan")]
+ res[indices,on=.(scan),precid:=i.precid]
+
+}
--
GitLab