Deep learning is a trending field in bioinformatics; so far, mostly known for image processing and speech recognition, but it also shows promising possibilities for data processing in food analysis, ...especially, foodomics. Thus, more and more deep learning approaches are used. This review presents an introduction into deep learning in the context of metabolomics and proteomics, focusing on the prediction of shelf-life, food authenticity, and food quality. Apart from the direct food-related applications, this review summarizes deep learning for peptide sequencing and its context to food analysis. The review’s focus further lays on MS (mass spectrometry)-based approaches. As a result of the constant development and improvement of analytical devices, as well as more complex holistic research questions, especially with the diverse and complex matrix food, there is a need for more effective methods for data processing. Deep learning might offer meeting this need and gives prospect to deal with the vast amount and complexity of data.
The quality of food is influenced by several factors during production and storage. When using marker compounds, different steps in the production chain, as well as during storage, can be monitored. ...This might enable an optimum prediction of food's shelf life and avoid food waste. Especially, proteoforms and peptides thereof can serve as indicators for exogenous influences. The development of a proteomics-based workflow for detecting and identifying differences in the proteome is complex and time-consuming. The aim of the study was to develop a fast and universal workflow with ultra-high temperature (UHT) milk as a proteinaceous model food with expectable changes in protein/peptide composition. To find an optimum shelf life without sticking to a theoretically fixed best-before date, new evaluation and analytical methods are needed. Consequently, a modeling approach was used to monitor the shelf life of the milk after it was treated thermally and stored. The different peptide profiles determined with high-resolution mass spectrometry (HRMS) showed a significant difference depending on the preparation method of the samples. Potential marker peptides were determined using orthogonal projections to latent structures discriminant analysis (OPLSDA) and principal component analysis (PCA) following a typical proteomics protocol with tryptic hydrolysis. An additional Python-based algorithm enabled the identification of eight potential tryptic marker peptides (with mass spectrometric structural indications
/
885.4843,
/
639.3500,
/
635.8622,
/
634.3570,
/
412.7191,
/
623.2967,
/
880.4767, and
/
692.4041), indicating the effect of the heat treatment. The developed workflow is flexible and can be easily adapted to different research questions in the field of peptide analysis. In particular, the process of feature identification can be carried out with significantly less effort than with conventional methods.
The analysis of almost holistic food profiles has developed considerably over the last years. This has also led to larger amounts of data and the ability to obtain more information about ...health-beneficial and adverse constituents in food than ever before. Especially in the field of proteomics, software is used for evaluation, and these do not provide specific approaches for unique monitoring questions. An additional and more comprehensive way of evaluation can be done with the programming language Python. It offers broad possibilities by a large ecosystem for mass spectrometric data analysis, but needs to be tailored for specific sets of features, the research questions behind. It also offers the applicability of various machine-learning approaches. The aim of the present study was to develop an algorithm for selecting and identifying potential marker peptides from mass spectrometric data. The workflow is divided into three steps: (I) feature engineering, (II) chemometric data analysis, and (III) feature identification. The first step is the transformation of the mass spectrometric data into a structure, which enables the application of existing data analysis packages in Python. The second step is the data analysis for selecting single features. These features are further processed in the third step, which is the feature identification. The data used exemplarily in this proof-of-principle approach was from a study on the influence of a heat treatment on the milk proteome/peptidome.
Graphical abstract
•Peptide hydrolyzates were characterized by two-dimensional thin-layer chromatography.•The repeatability of two-dimensional thin-layer chromatography was evaluated.•51 of 85 peptides were identified ...by MALDI-TOF-MS/MS.•Correlations between peptide properties and their retention factors were verified.•Retention factors based on amino acid sequence of the peptides were predicted.
High-performance thin-layer chromatography (HPTLC) is a suitable method for the analysis of peptides and proteins due to a wide selection of stationary and mobile phases and various detection options. Especially, two-dimensional HPTLC (2D-HPTLC) enables a higher resolution compared to one-dimensional HPTLC in the separation of complex peptide mixtures. Similar to 2D electrophoresis, characteristic peptide patterns can be obtained, allowing a differentiation of ingredients based on varying protein origins.
The aim of this study was to evaluate 2D-HPTLC with regard to its suitability for the characterization of proteins/peptides and to verify whether it is possible to predict the retention behavior of peptides based on their properties. As models, the five most abundant milk proteins α-lactalbumin, β-lactoglobulin, α-, β-, and κ-Casein were used. In order to determine the repeatability of the peptide separation by 2D-HPTLC, each tryptic protein hydrolyzate was separated eight times. The standard deviations of the retardation factors for the separated peptides varied between 1.0 and 11.1 mm for the x-coordinate and 0.5–7.3 mm for the y-coordinate. It was also shown that after the chromatographic separation, peptides of the individual protein hydrolyzates were located in specific areas on the HPTLC plate, so that a clustering could be obtained for the whey proteins‘ as well as the caseins‘ hydrolyzates. For establishing correlations between the properties of the peptides and their retardation factors, 51 of 85 selected peptides were identified by matrix-assisted laser desorption/ionization time-of-flight tandem mass spectrometry (MALDI-TOF-MS/MS). On this basis, statistically significant correlations (α = 0.05) between the retardation factors of the peptides and their isoelectric points, as well as the percentage of anionic and non-polar amino acids in the peptides were established. Finally, it was investigated, whether the retardation factors for peptides can be predicted on the basis of a linear regression of the percentage of non-polar amino acids in a peptide. For this purpose, a mixture of artifical (synthetic) peptides (n = 14) was separated by 2D-HPTLC and the measured retardation factors were compared with the corresponding retardation factors calculated. Absolute deviations of 0.3–17.9 mm were obtained. In addition, the universal applicability of the method to other protein sources other than milk proteins (animal protein) was tested using a mixture of pea peptides (plant protein, n = 3) resulting in absolute deviations of 0.7–8.6 mm.