Managing samples

Experiment designs are sometimes complex and may include:

• pooling samples together,
• processing samples with different protocols,
• processing samples using different assay types (e.g. performing multi-omics experiments including genomics, transcriptomics, proteomics, etc.)

Accurate experimental design representation, including sample relationships and sample lineages is important. LabID not only enables recording these complex relationships and lineage digitally (incl. split & merge) but also provide interfaces to examine, navigate, annotate and edit these samples and relationships.

Editing sample relationships

Samples can be created, individually or in batch. Relationships of samples can be edited manually on the sample detail page.

On the other hand, samples are often created and/or consumed within specific contexts. A sample can for example derive from (an)other sample(s), or be consumed, and/or created, within experiments. Creating samples capturing relevant lineage information is also possible.

• On the Specimen detail page:

  A sample can be derived from a parent specimen. The new sample will list its parent specimen.

• On the Experiment detail page, with the Sample Editor

  Experiment can consume or create samples (e.g. generating a sequencing library in the lab by extracting genomic DNA from cells). When editing an experiment, the detail page displays the sample editor panel at the bottom. This allows to batch create samples in the context of this experiment, meaning this experiment is becomes part of the lineage.

• When registering the raw datasets of an assay

  When registering raw datasets generated by an assay, samples are linked to the datasets. This is because every raw dataset is obtained from particular sample, and it is very important piece of information to capture for reproducibility. In this situation, one can either generate appropriate samples (whose information can later be updated by the user), or link already existing samples.

• When batch creating samples

  Parent samples for each new sample can be indicated for within the batch creation template.

Batch creating children samples

It is possible to batch create children samples, however there is currently no automated way to obtain a template with parent information pre-filled.

Currently, the parent sample has to be listed as a property of the child, using its UUID:

1. Use the sample export function to gather all parent samples UUIDs
2. Refer to these UUIDs when creating the children samples.

Search & Find samples

In order to find all samples associated to a project, you should use the context filter, instead of standard list filters. To set a context please refer to List views - Search & filter - Context

There are two kind of relationship between Samples and Projects

As detailed on the Sample lineage page, samples have two types of relationship with a project:

1. Relationship to the main project
2. Relationships to other projects through studies Setting Standard list filters with a project will only return samples for which the given project is the main project (as part of which they have been originally created).

Merge redundant samples

Sample duplication refers to having 2 sample items registered when it is in reality the same and unique entity.

Example of situations where a sample can get duplicated

Upon registering raw datasets, each dataset has to be associated to a sample. Samples can at this point be generated automatically. Albeit generally useful, certain situations instead require users to link again a sample that already exists in the system. This is true for example when:

• Re-sequencing

  the same library to gain additional sequencing coverage (2 datasets are obtained from a single library/sample)

• Pre-registration

  Samples have been registered by the user early on, prior to dataset registration, for example using the sample editor of the ELN.

Sample that have wrongfully been duplicated should be merged back together as soon as possible in order to keep the lineage information relevant.

The Merge Samples feature is accessible from the sample list page. The Merge Samples button is enabled when (1) multiple samples are selected and (2) the user has Edit & Delete permissions are available on all the selected samples

Technical overview of sample merging

A single sample is kept, the other ones are deleted. Relationships from & to removed samples are transferred to the kept sample

Samples can be automatically grouped into different groups using the Group By Name button. In this case, one sample is kept per group. When a sample does not group with others, it will be kept untouched.

Annotate samples

Annotations reflect project specific aspects of your samples. Enriching your samples with extensive annotations is key to benefiting from LabID and similar FAIR systems. Plus it is good practice, compliant to open science principles and mandatory when you want to publish data.

When samples are annotated, datasets and samples can be easily searched and discovered.

You can either:

• annotate individual samples from their edit page; or
• batch-edit them from the list page; or
• use the excel-based batch edition.