Skip to contents

FragPipe TMT QC

Functional Genomics Center Zurich

Source: vignettes/QCReport.qmd

Configuration

This section shows the key parameters used for this QC analysis. The report can be parameterized to analyze different datasets by providing custom file paths and identifiers.

project information
name value
workunit_id 1234
project_id 4321
datasetname ptm_example-main/qc_example_data/QCmini/psm.tsv
fastasequence ptm_example-main/qc_example_data/fgcz_3702_UP000006548_AraUniprot_1spg_d_20231024.fasta

QC will only use the first psm file.

using : ptm_example-main/qc_example_data/QCmini/psm.tsv

FASTA Database Summary

The FASTA database contains the protein sequences used for peptide identification. Understanding the composition of your search database is crucial for interpreting identification results and assessing potential biases.

Key metrics to evaluate:

  • Database size: Typical sizes vary by organism:
    • Human: ~20,000 proteins
    • Mouse: ~22,000 proteins
    • Yeast: ~6,000 proteins
    • E. coli: ~4,400 proteins
    • Arabidopsis: ~27,000 protein coding genes, ~35,000 proteins
  • Decoy sequences: Reverse sequences (REV_) are used to estimate false discovery rates
  • Amino acid composition: Should reflect the expected composition of your sample organism

The FASTA database has 55907 sequences including decoys, and 27953 without decoys. The amino acid frequency distribution below shows the composition of your search database, which should be consistent with the expected proteome composition.

nr sequences:
 27953
 length summary:
    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.
     5.0   198.0   348.0   404.8   520.0  5400.0
 AA frequencies:
      [,1]
 A  712383
 B       5
 C  213335
 D  608123
 E  760476
 F  484702
 G  730135
 H  256377
 I  601507
 K  720098
 L 1078566
 M  277709
 N  498002
 P  543057
 Q  395258
 R  609817
 S 1031163
 T  577392
 V  754323
 W  140645
 X      15
 Y  321973
 Z       3

Data Processing Overview

This analysis processes peptide-spectrum matches (PSMs) from FragPipe with specific quality filters to ensure reliable quantification.

Quality filters applied:

  • Peptide Prophet probability > 0.9: Ensures high confidence identifications
  • Abundance threshold > 0: Removes PSMs without quantitative information
  • Purity threshold = 0: No isolation purity filtering (all PSMs included)
Rows: 5436 Columns: 55
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr (14): Spectrum, Spectrum File, Peptide, Modified Peptide, Prev AA, Next ...
dbl (38): Peptide Length, Charge, Retention, Observed Mass, Calibrated Obser...
lgl  (3): Observed Modifications, Is Unique, Quan Usage

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
Rows: 5436 Columns: 55
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr (14): Spectrum, Spectrum File, Peptide, Modified Peptide, Prev AA, Next ...
dbl (38): Peptide Length, Charge, Retention, Observed Mass, Calibrated Obser...
lgl  (3): Observed Modifications, Is Unique, Quan Usage

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

For this analysis we are using all PSM (Spectra) reported in the psm.tsv file with a peptide prophet probability greater than 0.90.9, and an abundance value in any of the channels greater then 00. No other filtering is enabled. This reduces the number of PSM from 5436 to 5275.

The reduction in PSM count reflects the stringency of our quality filters. A typical reduction of 1-10%\% is expected and indicates proper quality control.

INFO [2025-12-14 21:01:32] get_annot : ptm_example-main/qc_example_data/fgcz_3702_UP000006548_AraUniprot_1spg_d_20231024.fasta
INFO [2025-12-14 21:01:33] get_annot : finished reading
INFO [2025-12-14 21:01:34] get_annot : extract headers
INFO [2025-12-14 21:01:34] get_annot : all seq : 55907
INFO [2025-12-14 21:01:34] removing decoy sequences usin patter : ^REV_|^rev_
INFO [2025-12-14 21:01:34] get_annot nr seq after decoy removal: 27954
INFO [2025-12-14 21:01:34] get_annot : isUniprot : TRUE
INFO [2025-12-14 21:01:34] get_annot : extracted gene names
INFO [2025-12-14 21:01:34] get_annot : protein length
INFO [2025-12-14 21:01:35] get_annot : nr of tryptic peptides per protein computed.
Warning in prolfquapp::dataset_protein_annot(psm, c(protein_Id = "Protein"), :
deprecated! use build_protein_annot
uniprot database : TRUE
Warning: Expected 2 pieces. Missing pieces filled with `NA` in 20 rows [54, 59, 92, 99,
178, 287, 288, 307, 357, 379, 475, 476, 554, 611, 754, 892, 893, 940, 967,
1021].
creating sampleName from fileName column
Warning in prolfqua::setup_analysis(psm, config): no isotopeLabel column
specified in the data, adding column isotopeLabel automatically and setting to
'light'.
Warning in prolfqua::setup_analysis(psm, config): no nr_children column
specified in the data, adding column nr_children and setting to 1.
completing cases
completing cases done
setup done

Identification Summary

These metrics provide an overview of the depth and breadth of your proteomic analysis. Higher numbers generally indicate better sample preparation and instrument performance.

Understanding the hierarchy:

  • Proteins: Unique protein groups identified
  • Peptides: Unique peptide sequences (without modifications)
  • Peptidoforms: Unique peptides with specific modifications
  • Precursors: Unique peptide-charge state combinations
  • PSMs: Total peptide-spectrum matches
Table: Nr of proteins, peptides, peptidoforms, precursors, spectrum peptide matches overall.
isotopeLabel protein_Id peptide_Id mod_peptide_Id precursor Spectrum
light 1064 3845 3947 4860 5275

The ratio between these levels can indicate data quality:

  • PSMs/Peptides ratio: Higher ratios suggest good reproducibility
  • Peptidoforms/Peptides ratio: Indicates the extent of post-translational modifications
  • Peptides/Proteins ratio: Reflects proteolytic efficiency and peptide diversity

Figure: Number of proteins, peptides, peptidoforms, ions and precursor per channel.

This plot shows the identification counts across different TMT channels. Ideally, all channels should show similar identification numbers, indicating consistent sample preparation and labeling efficiency.

Table: Nr of proteins, peptides, peptidoforms, precursors, spectrum peptide matches per channel.
isotopeLabel sampleName protein_Id peptide_Id mod_peptide_Id precursor Spectrum
light QCmini_126 1064 3845 3947 4859 5274
light QCmini_127C 1064 3845 3947 4859 5274
light QCmini_127N 1064 3845 3947 4860 5274
light QCmini_128C 1064 3845 3947 4860 5275
light QCmini_128N 1064 3845 3947 4859 5274
light QCmini_129C 1064 3845 3947 4859 5274
light QCmini_129N 1064 3845 3947 4860 5275
light QCmini_130C 1064 3845 3947 4859 5274
light QCmini_130N 1064 3845 3947 4860 5274
light QCmini_131C 1064 3844 3946 4858 5272
light QCmini_131N 1064 3845 3947 4859 5273
light QCmini_132C 1064 3845 3947 4859 5273
light QCmini_132N 1064 3845 3947 4860 5274
light QCmini_133C 1064 3845 3947 4859 5274
light QCmini_133N 1064 3845 3947 4859 5274
light QCmini_134C 3 3 3 3 3
light QCmini_134N 1064 3845 3947 4859 5274
light QCmini_135N 7 9 9 9 9

Quality indicators:

  • Consistent counts across channels: Good sample preparation
  • Large variations: May indicate labeling issues or sample degradation
  • Outlier channels: Should be investigated for technical problems

Modifications Summary

Post-translational modifications (PTMs) and chemical modifications provide insights into sample treatment and biological state. TMT labeling introduces specific modifications that should be monitored for labeling efficiency.

Key modifications to monitor:

  • TMT labels: N-term(229.1629), K(229.1629) for TMT10plex
  • Oxidation: M(15.9949) - common artifact
  • Carbamylation: N-term/K modifications from urea treatment
Number and type of modifications observed in the data.
Var1 Freq
C(57.0214) 2
K(304.2071) 2339
M(15.9949) 635
N-term(304.2071) 3904
N-term(42.0106) 10

Number and type of modification observed in the data.

The modification landscape reflects both intentional chemical treatments (TMT labeling) and unintentional modifications (oxidation, deamidation). High numbers of TMT modifications indicate successful labeling.

Labelling Efficiency

TMT labeling efficiency is critical for accurate quantification. Incomplete labeling leads to quantitative errors and reduced dynamic range.

Monitoring labeling efficiency:

  • N-terminal labeling: Should approach 100% for most peptides
  • Lysine labeling: Should be >95% for high-quality samples
  • Unlabeled peptides: May co-elute and interfere with quantification

N-term

N-terminal labeling efficiency is typically very high (>95%) as it’s kinetically favored.

Peptides

  • Total number of peptides : 3947
  • Number of peptides with modified N-term : 3904
  • Percent peptides with modified N-term: 99 %

Expected range: 95-100%. Values below 95% may indicate:

  • Suboptimal labeling conditions
  • Sample degradation
  • Buffer incompatibility

PSM’s

  • Total number of PSMs : 5275
  • Number of PSMs with modified N-term : 5227
  • Percent PSMs with modified N-term: 99 %

PSM-level statistics weight peptides by their identification frequency, providing insight into the most abundant species in your sample.

Lysine Modification

Lysine labeling is more challenging than N-terminal labeling and is more sensitive to reaction conditions.

Peptides 

Warning: Use of .data in tidyselect expressions was deprecated in tidyselect 1.2.0.
ℹ Please use `"modSeq"` instead of `.data$modSeq`
  • Total number of peptides with Lysine: 2339
  • Number of peptides with modified Lysine residues : 2339
  • Percent peptides with modified Lysine residues: 100 %

Expected range: 95-99%. Lower values may indicate:

  • Insufficient reagent concentration
  • Competing reactions (e.g., formaldehyde crosslinking)
  • pH optimization needed

PSM’s with Lysine

  • Total number of PSMs with Lysine: 3200
  • Number of PSMs with modified Lysine residues : 3200
  • Percent PSMs with modified Lysine residues: 100 %
Warning: Use of .data in tidyselect expressions was deprecated in tidyselect 1.2.0.
ℹ Please use `"n"` instead of `.data$n`
Warning: Use of .data in tidyselect expressions was deprecated in tidyselect 1.2.0.
ℹ Please use `"n1"` instead of `.data$n1`

Peptidoforms

  • Total number of Lysine residues: 2344
  • Number of modified Lysine residues : 2339
  • Percent modified Lysine residues: 100 %

Total number of Lysine residues in PSM’s

  • Total number of Lysine residues: 3205
  • Number of modified Lysine residues : 3200
  • Percent modified Lysine residues when taking number of PSMs into account: 100 %

Quantitative information per channel

TMT quantification relies on consistent labeling and equal loading across channels. These plots help identify technical issues that could bias quantitative comparisons.

Quality indicators:

  • Similar total abundances: Good sample preparation and loading
  • Outlier channels: May indicate pipetting errors or sample loss
  • Systematic patterns: Could suggest batch effects or labeling issues

Total abundance per channel (Sum of all abundances).

Total abundance reflects both sample amount and ionization efficiency. Large variations (>2-fold) should be investigated.

Ralative to chanell 126 total abundance per chanel.

Normalization to a reference channel (126) helps identify systematic biases. Values should typically be within 0.5-2.0 fold of the reference.

Warning: `aes_string()` was deprecated in ggplot2 3.0.0.
ℹ Please use tidy evaluation idioms with `aes()`.
ℹ See also `vignette("ggplot2-in-packages")` for more information.
ℹ The deprecated feature was likely used in the prolfqua package.
  Please report the issue at <https://github.com/wolski/prolfqua/issues>.

Density of abundance values per channel.

The abundance distribution should be similar across channels. Shifted distributions may indicate: - Unequal sample loading - Different sample complexity - Technical artifacts in specific channels

Missed cleavage

Missed cleavage analysis provides insights into proteolytic efficiency and sample quality. Excessive missed cleavages can reduce identification rates and affect quantification accuracy.

Missed cleavage site: Is a residue after which trypsin should have cleaved but did not.

Factors affecting missed cleavages:

  • Protein denaturation: Incomplete unfolding reduces accessibility
  • Digestion time/temperature: Insufficient conditions lead to incomplete digestion
  • Enzyme activity: Old or inhibited trypsin reduces efficiency
  • Chemical modifications: Modified residues may resist cleavage

To determine the number of missed cleavages we compute: - The number of all potential cleavage sites (i.e., number of K or R residues) - The number of actual cleavage sites (K or R at peptide C-terminus) - Modified residues that may resist cleavage

Missed Lysine residues

We compute the total number of K residues, and the number of K cleavage sites (nr of potential cleavage sites). Then we compute the number of K at the C-term

  • The number of K residues 3205
  • The number of unmodified K at C term 0
  • The number of modified K at C term 3025
  • The number of any K at C term 3025
  • The number of any modified K: 3200
  • Missed cleavage sites : number of K residues - number of any K at C term = 180,
  • and in % of number of K residues : 6

Expected range: 5-15% missed cleavages are typical. Higher rates may indicate:

  • Incomplete protein denaturation
  • Insufficient digestion time
  • Trypsin inhibitors present
  • High protein concentration

Missed Arginine residues

  • The number of R residues 2401
  • The number of unmodified R at C term 2189
  • The number of modified R at C term 0
  • The number of any R at C term 2189
  • The number of any modified R: 0
  • Missed cleavage sites : number of R residues - number of any R at C term = 212,
  • and in % of number of R residues : 9

Arginine cleavage is generally more efficient than lysine cleavage. Similar missed cleavage rates between K and R suggest consistent digestion conditions.

Session Info

R version 4.5.2 (2025-10-31)

Platform: aarch64-apple-darwin20

locale: en_US.UTF-8||en_US.UTF-8||en_US.UTF-8||C||en_US.UTF-8||en_US.UTF-8

attached base packages: stats, graphics, grDevices, utils, datasets, methods and base

other attached packages: here(v.1.0.2), ggseqlogo(v.0.2), lubridate(v.1.9.4), forcats(v.1.0.1), stringr(v.1.6.0), dplyr(v.1.1.4), purrr(v.1.2.0), readr(v.2.1.6), tidyr(v.1.3.1), tibble(v.3.3.0), ggplot2(v.4.0.1), tidyverse(v.2.0.0), prophosqua(v.0.1.0) and prolfquapp(v.0.1.8)

loaded via a namespace (and not attached): AhoCorasickTrie(v.0.1.3), gtable(v.0.3.6), xfun(v.0.54), prozor(v.0.3.1), htmlwidgets(v.1.6.4), ggrepel(v.0.9.6), lattice(v.0.22-7), tzdb(v.0.5.0), vctrs(v.0.6.5), tools(v.4.5.2), generics(v.0.1.4), parallel(v.4.5.2), pkgconfig(v.2.0.3), pheatmap(v.1.0.13), Matrix(v.1.7-4), data.table(v.1.17.8), RColorBrewer(v.1.1-3), S7(v.0.2.1), lifecycle(v.1.0.4), prolfqua(v.1.4.0), compiler(v.4.5.2), farver(v.2.1.2), codetools(v.0.2-20), htmltools(v.0.5.9), lazyeval(v.0.2.2), yaml(v.2.3.11), plotly(v.4.11.0), pillar(v.1.11.1), crayon(v.1.5.3), seqinr(v.4.2-36), MASS(v.7.3-65), cachem(v.1.1.0), tidyselect(v.1.2.1), conflicted(v.1.2.0), digest(v.0.6.39), stringi(v.1.8.7), pander(v.0.6.6), labeling(v.0.4.3), ade4(v.1.7-23), rprojroot(v.2.1.1), fastmap(v.1.2.0), grid(v.4.5.2), cli(v.3.6.5), logger(v.0.4.1), magrittr(v.2.0.4), patchwork(v.1.3.2), withr(v.3.0.2), scales(v.1.4.0), bit64(v.4.6.0-1), timechange(v.0.3.0), httr(v.1.4.7), rmarkdown(v.2.30), bit(v.4.6.0), gridExtra(v.2.3), hms(v.1.1.4), memoise(v.2.0.1), evaluate(v.1.0.5), knitr(v.1.50), viridisLite(v.0.4.2), dtplyr(v.1.3.2), rlang(v.1.1.6), Rcpp(v.1.1.0), docopt(v.0.7.2), glue(v.1.8.0), vroom(v.1.6.7), jsonlite(v.2.0.0) and R6(v.2.6.1)