<h1>BioCompute Object Project</h1>
<p><strong><em>*Links to Project Website and Resources:</em></strong>*</p>
<p>Project Website: <a href="http://biocomputeobject.org" rel="nofollow">http://biocomputeobject.org</a></p>
<p>BioCompute Object Specification Document V 1.0</p>
<p><a href="https://github.com/biocompute-objects/BCO_Spec_V1.0/blob/master/BCO_STU_v1.0.pdf" rel="nofollow">https://github.com/biocompute-objects/BCO_Spec_V1.0/blob/master/BCO_STU_v1.0.pdf</a></p>
<p>BioCompute Object Specification Document DRAFT V 1.2 </p>
<p><a href="https://docs.google.com/document/d/1w7PxAHbv5C-VNbLfelIQUFa6jDxnocaTjDwwUuG0eeY" rel="nofollow">https://docs.google.com/document/d/1w7PxAHbv5C-VNbLfelIQUFa6jDxnocaTjDwwUuG0eeY</a></p>
<p>BCO Proof of Concept (PoC) meeting minutes</p>
<p><a href="https://docs.google.com/document/d/173mh2kNn5AHD6rsA5KDxuyxEVIDA-27vYF-MM5rfVeQ/edit" rel="nofollow">https://docs.google.com/document/d/173mh2kNn5AHD6rsA5KDxuyxEVIDA-27vYF-MM5rfVeQ/edit</a></p>
A Framework for Communicating HTS Computational Analysis</p>
<p><strong>BioCompute Object Project Mission</strong></p>
<p>The goal of the BioCompute Working Group is to create a standard for documenting computational and provenance information for High-Throughput Sequencing workflows in a common structure that can be used to facilitate FDA regulatory submission, the BioCompute Object.</p>
<p>The BioCompute Object aims to facilitate HTS workflow exchange between the FDA, pharmaceutical companies, bioinformatics platform providers and researchers through the creation of regulatory submissions proof of concepts. </p>
<p><strong>High-Throughput Sequencing Analysis Issues</strong> </p>
<p>The Current focus of the BioCompute project is standardizing High-throughput sequencing (HTS) analysis communication. The two main issues associated with HTS analysis are unpredictability and lack of interoperability. We propose to address these issue through the use of BioCompute Objects (BCOs).
A physical experiment (physics, chemistry, biology, etc.) becomes more unpredictable with more variables. For this reason, scientists will record every variable in the protocol and try to control them. An in-silico experiment is no different. Computational biology algorithms can be just as volatile as a real-world physical experiment, but this fact is often overlooked.
Compounding this issue is the fact that no standardization or widely accepted schemas exist for recording the protocol, procedures, or metadata associated with the computational analysis. Often it can be near impossible to reproduce the results of an in-silico experiment because information on parameters and specific settings are missing or incomplete.</p>
<p>A BioCompute object is an aggregate of fields organized into seven domains:
1) <strong>identification and provenance domain</strong> - this domain captures information that makes a specific BCO easy to search and identify through a unique object ID assigned for each BCO and also a name field which contains standard biology terminology and terms from curated databases.
2) <strong>usability domain</strong> - provides specific description for what the BCO can be used for.
3) <strong>description domain</strong> - describes the external references, the pipeline steps and the relationship of the input/output domain.
4) <strong>parametric domain</strong> - customized parameters for the computational flow which can affect the output of the calculations and final analysis.
5) <strong>execution domain</strong> - fields required to execute the BCO in the portals.
6) <strong>input/output domain</strong> - these are the input files used in the BCO and what the output files entailed in terms of format and content.
7) <strong>error domain</strong> - contains two subdomains which reports empirical errors and algorithmic errors. </p>
<p>Three of the seven domains in a BioCompute Object are immutable upon assignment of the digital signature: 1) the Parametric Domain, 2) the Execution Domain and 3) the I/O Domain. </p>
<p><strong>Proof of Concept</strong></p>
<p>We are developing two ways to construct a BCO: directly using web forms designed to create and edit BCOs or on a software platform, like HIVE, Galaxy, Seven Bridges.
Proof of Concept (PoCs) can be based on real use cases or can be example use cases that mimic potential use cases. Some potential use cases:
- Pharma is planning on submitting to FDA NGS analysis results in the future. They can create BCOs using example data (synthetic or public data)
- The vendor that provides analysis to the pharma or any other industry would like to create BCOs to provide their pipelines and workflows in a standardized format so that they can be evaluated by SMEs within the company.
<strong>Join the BioCompute Working Group!</strong></p>
<p>We welcome participation from stakeholders interested in facilitating workflow exchange with the FDA. Please contact us at <a href="http://firstname.lastname@example.org" rel="nofollow">email@example.com</a></p>