Journal:MendeLIMS: A web-based laboratory information management system for clinical genome sequencing

From LIMSWiki
Revision as of 17:42, 9 December 2015 by Shawndouglas (talk | contribs) (Added content. Saving and adding more.)
Jump to navigationJump to search
Full article title MendeLIMS: A web-based laboratory information management system for clinical genome sequencing
Journal BMC Bioinformatics
Author(s) Grimes, Susan M.; Ji, Hanlee P.
Author affiliation(s) Stanford University
Primary contact Email: genomics_ji@stanford.edu
Year published 2014
Volume and issue 15
Page(s) 290
DOI 10.1186/1471-2105-15-290
ISSN 1471-2105
Distribution license Creative Commons Attribution 2.0 Generic
Website http://www.biomedcentral.com/1471-2105/15/290
Download http://www.biomedcentral.com/content/pdf/1471-2105-15-290.pdf (PDF)

Abstract

Background: Large clinical genomics studies using next generation DNA sequencing require the ability to select and track samples from a large population of patients through many experimental steps. With the number of clinical genome sequencing studies increasing, it is critical to maintain adequate laboratory information management systems to manage the thousands of patient samples that are subject to this type of genetic analysis.

Results: To meet the needs of clinical population studies using genome sequencing, we developed a web-based laboratory information management system (LIMS) with a flexible configuration that is adaptable to continuously evolving experimental protocols of next generation DNA sequencing technologies. Our system is referred to as MendeLIMS, is easily implemented with open source tools and is also highly configurable and extensible. MendeLIMS has been invaluable in the management of our clinical genome sequencing studies.

Conclusions: We maintain a publicly available demonstration version of the application for evaluation purposes at http://mendelims.stanford.edu. MendeLIMS is programmed in Ruby on Rails (RoR) and accesses data stored in SQL-compliant relational databases. Software is freely available for non-commercial use at http://dna-discovery.stanford.edu/software/mendelims/.

Keywords: Next generation sequencing; Clinical studies; Laboratory information management; Pathology; Genomics; Genetics

Background

With next generation DNA sequencing (NGS) now being a commonly adopted technology, the genetic analysis of large clinical populations has become practical and is widely used for identifying disease-related germline and somatic variants such as cancer mutations. The genetic variation from thousands of individuals can now be identified with NGS whole genome, exome, targeted and other resequencing approaches. Due to the dramatic increase in the number of NGS clinical genomics studies, it has become increasingly important to develop adequate laboratory information managements systems (LIMS) to manage the thousands of patient samples that are subject to NGS analysis. Tracking and managing the clinical sample workflow involved in NGS analysis is an extremely difficult task, given the logistical issues of enrolling patients, fragmented procedures for acquisition of clinical study samples, complex molecular preparation steps and the intricacies of the NGS processing pipeline. Commercial systems are available but typically are high cost and require extensive modification to address the specific needs of biomedical research groups conducting genetic analysis on populations.

As a general and unique solution to the needs of managing the experimental workflow for clinical genome sequencing projects, we developed MendeLIMS, a web-based, robust and flexible solution for integrating the management of clinical study samples and NGS processes. With respect to genetic studies, MendeLIMS functionality can be grouped into four major categories: (i) enrollment of patients and acquisition of clinical study samples, (ii) sample assessment and processing, (iii) genomic analysis through preparation of next generation DNA sequencing libraries or other molecular assays such as microarrays and finally, (iv) DNA sequencing of samples with associated quality control metrics. Tracking of sequencing steps is currently supported for the following Illumina NGS instruments: GAIIx, MiSeq, HiSeq, HiSeq2500, NextSeq but can easily be configured for any type of NGS instrument which follows a sequencing library to flow cell workflow. We maintain a publicly available demonstration version of the application for evaluation purposes at http://mendelims.stanford.edu.

Implementation

MendeLIMS is written in Ruby using the open source web application framework Ruby on Rails (RoR) and implementation is platform-independent. Instructions for installation are provided in Additional file 1. For our own in-house instance of MendeLIMS, our servers run Linux/Ubuntu and we use the MySQL relational database management system (RDBMS). The application is easily configured to use any other SQL RDBMS supported by RoR. Figure 1 shows a simplified database schema for the major tables. A more comprehensive schema is provided in Additional file 2.

Additional file 1. An installation guide for MendeLIMS.
Format: PDF; Size: 1,016KB Download file

Fig1 Grimes BMCBioinformatics2014 15.jpg

Figure 1: Database schema for MendeLIMS. Main entities and their relationships are shown in this diagram,
and a complete schema showing other ancillary tables is provided in the supplementary material.

Additional file 2. MendeLIMS database schema diagram.
Format: PDF; Size: 54KB Download file

The web interface is designed to handle a variety of queries in a modular format (Figure 2). To facilitate consistent data entry, MendeLIMS uses drop-down lists for seamless data validation whenever possible. The drop-down lists themselves are user-configurable by users with the appropriate authorization. Examples of user-configurable items include sample types, sequencing library multiplexing schemes, alignment references and DNA sequencers. All of the features are described in the user’s manual (Additional file 3).

Fig2 Grimes BMCBioinformatics2014 15.jpg

Figure 2: Query web interfaces for MendeLIMS. Database queries are managed by a series of web pages that
have a modular format. Different search parameters are included with drop down menus used for standardized search
terminology. Based on the needs of any given group, the search interface can be easily modified to accommodate
new search or entry functions.

Additional file 3. User’s guide for MendeLIMS.
Format: PDF; Size: 338KB Download file

The look and feel of the application is easily changed or customized since all web pages inherit styles from an application-wide cascading style sheet (CSS), and in keeping with RoR convention, overall page layout and navigation is controlled by a single HTML layout file.

Sample nomenclature

To enable accurate tracking of samples from their initial acquisition, through all key intermediate steps and ultimately to DNA sequencing, we developed a sample labeling nomenclature loosely based on the scheme employed by the Cancer Genome Atlas (TCGA) project (https://wiki.nci.nih.gov/display/TCGA/Working+with+TCGA+Data). We maintain the original unique sample barcode, and add successive suffixes to indicate processing performed.

Acquisition of clinical study samples

After enrollment into a study, patient samples and their characteristics are entered into MendeLIMS through a web interface (Figure 2). A unique identifier (ID) is assigned for each new clinical sample. The user has the option of entering sample-relevant clinical data including pathology information from clinical reports, digital images originating from pathology slides and general clinical information about the patient (Figure 3). For efficient subsequent retrieval of the physical samples, the storage freezer and container location is entered using a standard nomenclature. If email triggers are configured, an email is automatically sent to an identified central coordinator and/or to a specified owner for the particular clinical trial giving details of any new sample entered into the system. The web interface enables sample entry to occur at any location thus facilitating sample entry by various researchers and clinical coordinators.

Fig3 Grimes BMCBioinformatics2014 15.jpg

Figure 3: Workflow of MendeLIMS. Multiple steps of the sample acquisition workflow for clinical studies are
fully integrated with next generation or genomic assay procedures. This allows one to trace the genomic
analysis of any given sample.

Clinical study sample assessment and processing

Any manipulation of clinical study samples is tracked (Figure 3). This includes dissection of tissue samples and nucleic acid extraction. Details of these sample workflow operations are stored in MendeLIMS including volumes of any extracted macromolecules such as genomic DNA, concentration metrics and sample storage location. This greatly facilitates the managements of these precious resources for population studies.

Tracking molecular and genomic analysis

Molecular assays and sequencing library steps are also captured in MendeLIMS (Figure 3). Sequencing libraries may be entered as singleplex (e.g. one sample per library), or multiplex (e.g. multiple samples per library with each sample tagged with a unique starting sequence). The multiplex indexing schemes are user-configurable, both for number of samples which can be multiplexed on one lane, and for the actual starting sequences used.

Tracking the next generation sequencing workflow

In preparation for initiation of an NGS analysis, a sequencing run is entered into the system by selecting existing libraries and placing them into separate lanes or partitions. Normally an entire sequencing run is entered. However, the system is also able to handle partial sequencing runs to accommodate the scenario where sequencing may be performed as a service and the run is shared between multiple groups who are not privy to each other’s results. MendeLIMS generates a unique sequencing run key based on the sequencing date, sequencing machine, and a unique sequential run number. Once the sequencing run has completed, the initial quality control (QC) metrics for the run can be entered into the system. This is currently handled by an offline ruby script, but will in future be incorporated into the web application. MendeLIMS supports any type of sequencing application including whole genome, exome, targeted and RNA-based sequencing studies. The system stores sequencing library and sample lineage, flow cell composition, and sequencing run metadata, along with run status and QC metrics for all runs (Figure 4). The sequencing data files — for example bam alignment files, or vcf variant calling files — are not stored in MendeLIMS per se but are on a storage cluster accessible to all researchers in the group. Additionally, since we use the MendeLIMS sequencing run key and sequencing library/sample nomenclature in the analysis directory and file names, the files are easily cross-referenced between MendeLIMS and the storage cluster.

Fig4 Grimes BMCBioinformatics2014 15.jpg

Figure 4: Tracking the sequencing of clinical samples. One can follow a clinical sample from enrollment in a clinical study all the way through
to its sequencing. For example, from the sequencing run composition one can back track to the individual libraries and the original source DNA.
Screen shots show the various levels of querying. A sequencing library can be queried for additional information. When required, it is possible to even
determine the time of enrollment in a study and pull up relevant images from pathology.

References

Notes

This presentation is faithful to the original, with only a few minor changes to presentation. In some cases important information was missing from the references, and that information was added.