6. Creating Projects and Collections

If possible, it is recommended that projects be created from within the collection they belong to, for reasons described in the following sections. Collections (and projects) can be created from Mosaic's landing page. Projects can also be created from within a collection.

6.1. Creating New Projects and Collections

A new project can be created inside a collection from Projects in the left menu. At the top of the table listing all the projects in the collection, click the + Create Subproject button. This will open the Create Project form. The different elements of this form are described in section 6.2 below.

On the main Projects page, click the + Projects button at the top of the table listing all available projects to open the create project form.

Tip: A project created inside a collection is automatically shared with all collection members at the role they hold on the collection, and any template and scopes of work associated with the collection are applied for you. Creating the project standalone and adding it to a collection afterwards means doing that sharing and setup by hand. Whenever a case belongs to a cohort, create it from within that cohort's collection.

6.2. The Create Project Form

Wherever the project is being created from, the Create Project form will need to be filled out describing the new project. The Name is required and should be informative.

Privacy Level

The Privacy Level is required and can take one of three values:

Public: the project will be visible and accessible to anyone.
Protected: users need to be granted access to the project. The Mosaic landing page provides a list of All Projects where users can see and request access to protected projects. Any requests will be sent to users with Admin status or above where they can grant access if desired.
Private: users need to be granted access to the project, but unlike protected projects, private projects are hidden and so users cannot find the project and request access.

The privacy level can be changed later from the project settings, so a project can be created privately and later made available more broadly. See Project Members for managing who has access, and Policies for restricting access to specific content within a project.

Project Type

When creating a project from within a collection, Collection of Projects will be disabled and the type will be set to Single Project or Subproject in Collection. When creating a project from Mosaic's landing page, the type can be set to Single Project or Collection of Projects.

Creating a collection: If Collection of Projects is chosen, the only additional information required is which projects should be added to the newly created collection. No projects have to be added, but the dropdown provides a list of available projects which can be selected. Any projects selected will be added to the collection when it is created.

Creating a project: If a project is created, more information is required. First a reference genome must be selected (see Reference Genome below).

Template: A dropdown menu provides a list of available templates. When the project is created, the selected template will be applied, importing attributes, creating conversations, and providing a standard look and feel for projects. See Templates.

Scope of Work: The scope of work defines which project attributes require values to be set. This can include attributes like the IRB number, project PI, and any other required information. Once a scope of work is selected, a form will appear listing all attributes in the scope of work, and those that are required are highlighted. Attributes in the scope of work can be given values in a template. When a template and scope of work are used together, scope-of-work attributes given values in the template will be automatically populated. Two scopes of work can be selected: an institutional scope of work is typically used for administration purposes, e.g. recording billing accounts etc. The program scope of work is used to define attributes common to similar projects. For example, if there exists a collection of neonatal sequencing projects, the program scope of work may contain attributes such as the name of the neonatologist assigned to the neonate.

The scopes of work ensure that users are prompted for necessary information on project creation, and the templates allow data to be automatically populated to minimize the time required to fill in these forms. If attributes are marked as required but are not given values when the project is created, a task will be created ensuring that after the project is created, the user is aware of outstanding tasks. See Tasks.

Reference Genome

When creating a project, a reference genome must be selected. Currently all data in a single project must come from the same reference genome, so choose the genome build that matches the sequencing data for the case. This is an important decision: variant coordinates, annotations, and the genes available for filtering are all tied to the project's reference genome, and it cannot be mixed within a project - and genome reference cannot later be changed for a project. If a project has data aligned to more than one genome build, the data for each build belongs in a separate project.

Family type

When a project is created, a preset family type can be selected, for example a trio. The Family name can be defined and the sample names changed, preferably to the IDs used in the sequencing data. These can be edited after project creation. Additional family members can be added by selecting Add a family member beneath the table of samples.

Tip: Setting the sample names to match the identifiers used in your sequencing data at creation time makes it easier to attribute data files to the correct samples later. For example, setting the proband sample ID in the Samples table to match the sample ID in the underlying VCF.

6.3. After Creating a Project

Creating the project is the first step. A typical case is then made ready for analysis by:

1. Creating all samples. Create all project samples and assign sample attributes (sex, relation, affected status, etc)
2. Confirming data files are linked. Check that the case's sequencing data is present and attributed to the correct samples under Settings & More > Data Files. See Data Files. Data loading is usually arranged with the Frameshift team during onboarding.
3. Recording the phenotype. Add the proband's HPO terms so that phenotype-driven variant prioritization is available.
4. Completing outstanding tasks. Fill in any required attributes flagged as tasks when the project was created.
5. Inviting collaborators. Add the team members who need access, at the appropriate roles (see Project Members). If the project was created inside a collection, members may already have access.

With those in place, the case is ready for variant analysis.

6.4. Recommended Workflow for a New Program

When standing up a new cohort, a little setup up front pays off across every case:

1. Create the collection for the cohort.
2. Work with the Frameshift team to create a template for the cohort and associate it with the collection (see Templates).
3. Define the scopes of work that capture the information every case must record.
4. Create each case as a subproject inside the collection, so the template and scopes of work are applied automatically and members are granted access without extra steps.

This way, the structure is defined once and every subsequent case inherits it.