Supplementary materials for paper: Computational challenges to reproducibility, robustness, extensibility and reuse in multi-omics: a meta-workflow-based case study (Under review)

---

This page provides pointers to the materials (datasets, software, workflows, provenance traces and container image) used in a paper currently under review (more details about the publication will be announced at a later stage)

Summary: The primary goal of our study is to demonstrate an example computational paradigm that uses meta-workflows (i.e., workflows that assess the results of two or more workflows) alongside abstract components (i.e, components that can be implemented with multiple tools) to overcome barriers facing multi-omic discovery. We characterize the reproducibility of multi-omics discoveries by attempting a figure-for-figure reproduction of the seminal paper by Zhang et al [1]. We then assess the robustness of its findings by re-running all the experiments with newly acquired data, discussing the differences of the obtained results.

This page and the materials described on it (excluding external references) is available in Zenodo under DOI 10.5281/zenodo.4633542.

This HTML page itself is accessible under the following permanent identifier: https://w3id.org/dgarijo/ro/zhang_et_al_2014_reanalysis_and_reuse

Datasets.

---

Input Datasets

Note that availability on some of the following datasets may be subject to the the platforms where they are stored:

• Proteomic Analysis of Colon and Rectal Carcinoma Using Standard and Customized Databases
• Custom and Standard reference sequence databases (including custom_merged_OV.fasta, custom_merged_BRCA.fasta, custom_merged_CRC95.fasta as custom datasets; and hg19_refseq_pro_seq.fasta, humanRefSeq_Version54_with_tryp_DECOY.fasta as reference datasets)
• MzML files corresponding to 95 samples were obtained from CPTAC
• RAW files for 41 samples from the NCI-60 full proteome data were downloaded from ProteomicsDB
• Proteomics data for 174 ovarian cancer samples was obtained from CPTAC
• 37 mzML files corresponding to 108 Breast cancer samples (including three duplicates and three normals from global proteome data)

Software and Workflows.

---

The pointers for using the main software and workflows can be found below:

• List of the software packages (and corresponding version) installed in the Docker image
• Documentation of the input and output files used by each workflow
• Documentation of the parameters used for each workflow
• List of the code wrappers used to create workflow components for all workflows. You can also browse them on GitHub. (Note: these wrappers do not contain the dependencies required for each tool, they are ready to use in the Docker image)
• Document with an overview of the workflows used in the manuscript. The workflows themselves can be found within the Docker image (in WINGS), or by downloading the workflow specification files (which contain all the implementation details). If you want to explore the contents of the zip files, you can also browse the on GitHub. (Note: data is not included in these files)

Docker image.

---

• A Docker image with WINGS and sample data, software and workflows is available on Zenodo
• The Docker image is currently available in DockerHub with the id: kcapd/wings-genomics-full:latest (availability may be subject to DockerHub policies)
• The instructions for downloading and running the Docker image locally can be found on GitHub.

Provenance traces.

---

Below you can find provenance traces for the experiments run in the paper, following the OPMW-PROV model (and extending the W3C PROV standard). Note that the pointers to code and data are relative URLs, which won't resolve (they indicate the names of the files that have been used, but the paths are local to the Docker image). The code components and data referred to can be found in the dataset and code sections.

• Proteomic Run (Stage 3 reanalysis).
• Proteogenomic Run.
• Pepitome Run (Figure 3).
• Xtandem Run (Figure 3).
• Myrimatch1: (Figure 3).
• Myrimatch2: (Figure 3).
• MSGFPlus: (Figure 3).

Bibliography.

---

[1] Zhang, B. et al. Proteogenomic characterization of human colon and rectal cancer. Nature 513, 382-387, doi:10.1038/nature13438 (2014)

About the authors.

---