About group analyses

Group analysis in proteomics refers to the systematic comparison of protein abundance across different biological samples or conditions. This process involves grouping samples based on specific criteria—such as treatment conditions, disease states, or genetic backgrounds—and analyzing the differences in protein abundance within these groups.

Group analysis enables researchers to uncover the molecular basis of biological processes and disease states, potentially leading to the identification of biomarkers and therapeutic targets.

When you set up a group analysis, you must specify:

• the data type: protein or peptide

• the statistical test: T test A statistical test used to determine if there is a significant difference between the means of two groups. It's commonly used when you have a continuous outcome variable (such as height, weight, etc.) and you want to compare the means of two independent groups (such as test scores of two different classes). A t-test is recommended when you have at least 30 samples in your group analysis. or Wilcoxon Rank-Sum Test The Wilcoxon rank-sum test, also known as the Mann-Whitney U test, is a non-parametric test used to compare two independent groups when the assumption of normality is violated or when the data is ordinal. A Wilcoxon statistical test is recommended when you have less than 30 samples in your group analysis.

• a category and the two groups within it to compare (more A category is a broad characteristic used to refine your analysis (such as sexes, age, treatment site) and the groups are specific values within the category (such as female or male under the category of sex).)

You can also specify optional settings including:

• filters to apply to sub-groups (more You may further refine your group analysis by filtering the dataset using other sample metadata or subgroups. For example, if you are comparing data annotated with both condition (healthy/disease) and age, you might want to filter the condition data for samples that appear only in one or more specific age subgroups. To filter using sample metadata, select the category that includes the subgroups you want to filter. Then select groups within that category to include in the group analysis. If no subgroups are selected, data from all subgroups will be included in the analysis. )

• individual samples to exclude

• quality thresholds where results that have undetected values by set number or percentage values can be excluded from the analysis (more Mass spectrometry analysis results may include missing entries, because not all proteins and peptides are consistently detected across samples. It's essential to address these missing values to ensure valid and useful statistical tests. One common method is to filter out proteins or peptides that don't meet quality thresholds. These thresholds, based on minimum detection levels or reliability criteria, help focus the analysis on more reliable data points, improving the accuracy and interpretability of the results. )

• a normalization method to reduce data variability that may arise from differences in sample preparation, instrumentation, or other technical factors: none log10 (unnormalized) Use for quick exploratory analysis or when data consistency is already well controlled. , MaxLFQ Best for large datasets with significant variations in protein abundance and when precise quantification is critical. , or median normalization Suitable for balancing overall intensity levels across samples in a straightforward manner, especially when dealing with moderate sample size and variability.

The Group Analysis Setup screen.

As you make selections, the Group Analysis Setup page displays real-time values, such as number of samples to be evaluated, based on the settings you've chosen.

The Group Analysis Summary section on the setup screen