What’s New in Omics Playground V3.5:
A Complete Guide to the Latest Features
Published on November 12th, 2024
⏱ 13 min read
Introduction
We are thrilled to announce the latest release of Omics Playground, bringing you powerful new tools and features to elevate your omics data analysis experience.
As our multi-omics community continues to grow, we recognize the need for solutions that are more personalized, intuitive, and scalable. With this release, we’ve enhanced the data upload module, added new proteomics-focused analysis options, and boosted platform performance—enabling you to derive deeper insights from your RNA-Seq and proteomics datasets with greater ease.
Additionally, we’re excited to share that metabolomics support and a dedicated upload board for single-cell RNA-seq data are on the horizon! While these features aren’t public just yet, we’re aiming a release for metabolomics support by the end of the year.
Whether you’re a seasoned Omics Playground user or just starting out, these updates are designed to empower your research and accelerate discoveries. Here’s a look at what’s new in this latest version:
New guided data upload module and proteomics-focused options
Ability to skip missing values imputation
New normalization methods for proteomics data
Outlier detection
Unsupervised batch correction
Ability to append protein symbols to feature IDs
Defaults for differential expression testing methods for proteomics data
Analysis at residue/peptide level for phosphoproteomics data
- Accept PTM data format (Uniprot + site position)
- Automatic residue annotation of phosphorylation site
- Added links to proteomics databases in protein info box
Expanded support to 290+ species
Summary report generation and enriched information on modules and plots
Improved visualization options
Volcano plots coloring and labels
Dynamic renaming of y-axis in plots for proteomics
Display nominal p-values.
Display barplot of number of detected features per sample (or group)
Read on to explore the exciting new features and improvements in detail!
1. New guided data upload module and proteomics-focused options
We’ve listened to your feedback and made data upload easier with a step-by-step guide for an error-free experience (Video 1). You can learn more about each step in our Omics Playground data upload tutorial.
Video 1. A quick walkthrough of the new guided data upload process in Omics Playground.
Along with a brand new module, we’ve introduced several new data upload features that are worth exploring especially if you’re working with proteomics data:
- Ability to skip missing values imputation
- New normalization methods for proteomics data
- Outlier detection
- Unsupervised batch correction
- Ability to append protein symbols to feature IDs
- Defaults for differential expression testing methods for proteomics data
1.1 Ability to skip missing values imputation
Handling of missing values (NAs) has been improved and the user can select whether to impute or keep missing values. How to handle missing values (impute or skip imputation), and whether treating zero as missing values or not can both be chosen at step 4 of the upload dialog by selecting one of the methods from the left (Figure 1).
- Missing values can be imputed.Current available algorithms: svdImpute.
- Imputation can be skipped. If desired, imputation can be skipped and missing values will be retained in the data as NA.
- Zero values can be regarded as NA. Zero values can be regarded as NA and either be imputed or left as NA.
1. 2 New normalization methods for proteomics data
We’ve introduced several new normalization methods which can be chosen at step 4 of the upload dialog by selecting one of the methods from the left (Figure 2).
- MaxMedian
Sample-wise median normalization where medians of all samples are set to the highest median among the samples. - MaxSum
Sample-wise total intensity normalization where the sums of all samples are set to the highest sum among samples. - Reference normalization
Normalization to a selected reference protein. This normalization is recommended for smaller datasets with few proteins where the median or sum may not be robust. This method requires a stable protein as reference.
- MaxMedian
1.3 Outlier detection
To aid with outlier detection, we implemented scoring samples with an “outlier scores” based on a combination of z-scores (correlation, euclidean distance and feature-wise z-scores). You’ll find this option in point 3 of the QC/BC step of the data upload (Figure 3).
1.4 Unsupervised batch correction
Proteomics data often suffers from batch effects that reduce the statistical power to detect the true biological signal. We have incorporated state-of-the-art batch effect correction methods into OmicsPlayground (Figure 4). Below you’ll find a brief explanation of each of these methods.
- Auto-Combat
This is an automatic implementation of the ComBat batch correction algorithm that uses all available phenotype columns that are not-correlated with the phenotype of interest. - RUV (‘remove unwanted variation’)
RUV is an unsupervised method to remove unwanted variation. We use “very insignificant” (highest p-value) features as control features. - SVA
Surrogate Variable Analysis is widely used as an unsupervised batch correction method. We use default settings for SVA. - NPM
Nearest-Neighbour Matching corrects batch effects by matching closest positive and control pairs. This is an in-house developed method which is currently submitted for publication. You can read more about it here.
- Auto-Combat
BigOmics Tip: Please be aware that the QC/BC settings selected during upload aren’t yet available for download after dataset computation.
To retain this information for future reference, we recommend adding a note in the dataset description when prompted before clicking “Compute.” For instance, you might name the dataset “dataset-SVA” and include “SVA batch corrected” in the description.
1.5 Ability to append protein symbols to feature IDs
This feature appends the symbol to the feature name (such as Uniprot ID) for easier reading of the labels. You can find this option under ⚙️ Options at the Compute! step of the data upload (Figure 5).
1.6 Defaults for differential expression testing methods for proteomics data
T-test and moderated limma t-tests are automatically set as the default methods for differential expression testing when datatype is proteomics. You see this option under ⚙️ Options at the Compute! step of the data upload (Figure 6).
2. Analysis at residue/peptide level for phosphoproteomics data
In addition to “classical” proteomics analysis, Omics Playground has been adapted to handle post-translational modifications (PTM) data. PTM data involve measurements of post-translational biochemical modifications at specific amino-acid residues at peptide level. Minimally, PTM data would include protein identifiers and abundance data. Typically, data may also include the amino-acid (“residue”). Our new features handle these omics in PTM data.
2.1 Accept PTM data format (Uniprot + site position)
Phospho-PTM data is recognized automatically by uploading a datasheet with Uniprot identifiers in the first column and phospho site position in the second column. These two columns will be automatically merged and used as unique identifiers. Multiple site positions are allowed. See an example of PTM data format accepted in Omics Playground in Figure 7.
2.2 Automatic residue annotation of phosphorylation site
Phosphorylation positions are automatically annotated when uploading Uniprot and site position. Note: you need to select “proteomics” as data type and upload the expression.csv file with Uniprot identifiers in the first column and site position in the second column (Figure 8).
2.3 Added links to Proteomics databases in Protein info box
For each detected gene symbol, all Uniprot IDs available in the data are listed in the protein info box. Each Uniprot id has a hyperlink to the UniProtKB database (Figure 9). Links to PhosphoELM for peptide/protein id and corresponding gene are also provided.
3. Expanded support to 290+ species
Omics Playground has recently expanded its species support, now accommodating 290+ species including most animals, plants, fungi and more.
If you’re not sure your species is included you can either search the list during your data upload (Figure 10) or contact our team. Let us know which species you’re interested in and we’ll let you know if it’s currently supported.
Users can also upload their own organism-specific gene sets at the last stage of the data upload process (Figure 11).
4. Summary report generation and enriched information on modules and plots
You can now download a summary report of a datasets in your dataset library by selecting a dataset and clicking generate a report (Figure 12).
You’ll be prompted to select the report type and see an example of how the report will look before clicking Download (Figure 13).
Additionally, we’ve enhanced the information available in each module, giving you more detailed insights into the methods used. This information can be copied and included in your publication if needed.
Finally, when downloading a plot as a PDF, you can now view the options and settings you selected to generate it, provided as supplementary information (Figure 14 and 15).
5. Improved visualization options
In addition to the enhancements mentioned above, we’ve made several other improvements to the platform. Here are some of the most noteworthy updates:
5.1 Volcano plots coloring and labels
Volcano plots now display up-regulated genes in red and down-regulated genes in blue, with the option to view them in either dynamic or static mode. You can easily switch between the two by selecting the desired option at the top of the plot.
Additionally, labels can be displayed by symbol or feature name. In feature level plots, the user can choose to label the points by symbol or by full feature name (Video 2).
Video 2. Updated coloring and labeling for volcano plots in Omics Playground.
Finally, the user is also now able to visualize just the top 10 up-regulated and down-regulated features in the differential gene expression table located below the volcano plots.
5.2 Dynamic renaming of y-axis in plots for proteomics
The y-axis of plots display “Abundance” or “Abundance (log2)” as a measure of protein expression (Figure 16).
5.3 Display nominal p-values
A new option “Show p-values” has been implemented to enable display of p-values in differential gene expression tables and volcano plots (Figure 17).
5.4 Display barplot of number of detected features per sample (or group)
You’ll now see an option to display the barplot in DataView>Sample QC by number of detected features (Figure 18).
Conclusion
Omics Playground is now fully optimized for advanced proteomics data analysis, enabling you to dive deeper into your research with ease.
As a final note, you can now explore metabolomics data analysis in beta mode, allowing you to test new features and workflows before they are fully released. To enable beta mode, simply navigate to the settings and toggle the option for beta features (Figure 19).
If you’d like to know more about the new features or want to share your experience with the new release, feel free to reach out to our team here. We’re always eager to hear your feedback, as it helps us improve your experience!
Don’t wait, explore the latest Omics Playground features today and see how they can enhance your omics research!
New to Omics Playground? Register below for a free trial!