Journal: Netherlands Heart Journal

Article Title: Data quality and auditing within the Netherlands Heart Registration: using the PCI registry as an example

doi: 10.1007/s12471-022-01752-1

Figure Lengend Snippet: Flowchart of the Netherlands Heart Registration ( NHR ) NEN-7510 certified data quality assurance system

Article Snippet: Within the NHR a Royal Netherlands Standardization Institute (NEN) 7510 certified data quality assurance system is used (Fig. ).

Techniques:

Journal: Scientific Data

Article Title: Direct infusion mass spectrometry metabolomics dataset: a benchmark for data processing and quality control

doi: 10.1038/sdata.2014.12

Figure Lengend Snippet: All samples were analysed in triplicate in positive ion mode, with each analytical batch comprising of 20 biological samples and 5 equivalent QC samples; extraction blanks (B) were also analysed. The analytical batch was measured 8 times, across 7 days, while instrumental factors associated with normal mass spectrometer use were changed between batches to assess their impact on analytical variability. Reprinted from Kirwan, Broadhurst et al. , Characterising and correcting batch variation in an automated direct infusion mass spectrometry (DIMS) metabolomics workflow, Analytical and Bioanalytical Chemistry 405(15): 5147–5157, with kind permission from Springer Science and Business Media.

Article Snippet: While a number of methods exist for the quality assurance and quality control of metabolomics data, their use is not yet widespread nor have these methods been developed into formal recommendations by relevant international organisations such as the Metabolomics Society.

Techniques: Extraction, Mass Spectrometry

Journal: Scientific Data

Article Title: Direct infusion mass spectrometry metabolomics dataset: a benchmark for data processing and quality control

doi: 10.1038/sdata.2014.12

Figure Lengend Snippet: Data processing workflow for direct infusion FT-ICR mass spectrometry-based metabolomics dataset. All cow (C), sheep (S) and Quality Control (QC) samples were analysed in triplicate (e.g., for cow 5, analytical replicates C5, C5’ and C5’’) using direct-infusion mass spectrometry (DIMS). Samples with visual evidence of a poor electrospray current or with an outlying total ion count (TIC) profile were flagged as technical outliers. Data and metadata in the instrument .RAW files (IRF) and averaged transients (AT) were subject to apodisation, zero-filling and fast Fourier transformation (FFT). Next, the resulting frequency spectra (FS) were mass calibrated and SIM-stitched, which resulted in three stitched peak lists (e.g., SPL C5 , SPL C5’ and SPL C5’’ ) for each sample. The peaks in each SPL were subject to four further filters in order to discard noise and retain only the more robust peaks, comprising of replicate filtering, blank filtering, sample filtering and missing-value filtering. These processing steps generated two datasets: a replicate filtered peak list (RFPL) for each sample, and a single sample filtered peak matrix (SFPM) for the whole study. The technical outliers as identified by the TIC-filtering were removed from the dataset immediately prior to sample filtering. The SFPM dataset was further processed using probabilistic quotient normalisation (PQN), QC-robust spline batch correction, and spectral cleaning. Finally, each of these three data matrices was subject to missing values imputation using KNN and then variance stabilisation using a generalised logarithm transformation. Light-gray boxes represent the different datasets, blue trapezoids represent the data filtering steps, and green ovals represent the remaining processing steps.

Article Snippet: While a number of methods exist for the quality assurance and quality control of metabolomics data, their use is not yet widespread nor have these methods been developed into formal recommendations by relevant international organisations such as the Metabolomics Society.

Techniques: Mass Spectrometry, Control, Transformation Assay, Generated

Journal: Scientific Data

Article Title: Direct infusion mass spectrometry metabolomics dataset: a benchmark for data processing and quality control

doi: 10.1038/sdata.2014.12

Figure Lengend Snippet: ( a ) Bar chart showing the percentage of missing values in the mass spectrum of each QC and biological sample. The black dashed line represents the mean number of missing values plus two standard deviations, and is used as a 95% exclusion threshold. One sample (QC35) clearly exceeds the acceptability threshold for the number of missing values and therefore was excluded from the dataset. ( b ) PCA scores plot of the DIMS metabolomics dataset SFPM PQN+BATCH+CLEAN+KNN+GLOG (PQN-normalised, batch-corrected, spectral cleaned, KNN missing value imputed and generalised logarithm transformed), colour-coded according to 20 biological samples and the QC samples. The tight clustering of the QC samples confirms that the analytical variation, both within and across all 8 batches, is small relative to the biological variation. Key: O sheep (multi-colours), ▽ cow (multi-colours), ◊ QC (red).

Article Snippet: While a number of methods exist for the quality assurance and quality control of metabolomics data, their use is not yet widespread nor have these methods been developed into formal recommendations by relevant international organisations such as the Metabolomics Society.

Techniques: Transformation Assay

Journal: Scientific Data

Article Title: Direct infusion mass spectrometry metabolomics dataset: a benchmark for data processing and quality control

doi: 10.1038/sdata.2014.12

Figure Lengend Snippet: Analytical precision of the DIMS metabolomics datasets presented as median RSD QC values (%).

Article Snippet: While a number of methods exist for the quality assurance and quality control of metabolomics data, their use is not yet widespread nor have these methods been developed into formal recommendations by relevant international organisations such as the Metabolomics Society.

Techniques:

Journal: Scientific Data

Article Title: Direct infusion mass spectrometry metabolomics dataset: a benchmark for data processing and quality control

doi: 10.1038/sdata.2014.12

Figure Lengend Snippet: Analytical precision of the DIMS metabolomics datasets presented as median RSD biol values (%).

Article Snippet: While a number of methods exist for the quality assurance and quality control of metabolomics data, their use is not yet widespread nor have these methods been developed into formal recommendations by relevant international organisations such as the Metabolomics Society.

Techniques: