A primer to clinical genome sequencing

Abstract

Purpose of review: Genome sequencing is now available as a clinical diagnostic test. There is a significant knowledge and translation gap for nongenetic specialists of the processes necessary to generate and interpret clinical genome sequencing. The purpose of this review is to provide a primer on contemporary clinical genome sequencing for nongenetic specialists describing the human genome project, current techniques and applications in genome sequencing, limitations of current technology, and techniques on the horizon.

Recent findings: As currently implemented, genome sequencing compares short pieces of an individual's genome with a reference sequence developed by the human genome project. Genome sequencing may be used for obtaining timely diagnostic information, cancer pharmacogenomics, or in clinical cases when previous genetic testing has not revealed a clear diagnosis. At present, the implementation of clinical genome sequencing is limited by the availability of clinicians qualified for interpretation, and current techniques in used clinical testing do not detect all types of genetic variation present in a single genome.

Summary: Clinicians considering a genetic diagnosis have wide array of testing choices which now includes genome sequencing. Although not a comprehensive test in its current form, genome sequencing offers more information than gene-panel or exome sequencing and has the potential to replace targeted single-gene or gene-panel testing in many clinical scenarios.

Figure 1. An Illustration of the Steps in Clincial Genome Sequencing

A. Generating Raw Sequencing Data: After obtaining informed consent from the patient and family which includes discussion of family preferences regarding incidental findings, a sample of DNA is obtained from the patient and subject to fragmentation and preparation into a “library” for sequencing. The library is subject to sequencing which yields raw data in the form of unordered “reads” of sequence between 50–250 base pairs in length. B. Mapping and Variant Calling: The raw data is mapped to the reference sequence from the human genome project resulting in an ordered alignment of the reads tiled across the entirety of the known reference sequence. In this representation of an alignment, matches to the reference sequence are denoted by dots while mismatches are represented by a capital letter. Some reads may fail to be aligned or alternately may be aligned incorrectly in the wrong location. When a mismatch between the reads and the reference genome is detected in enough reads to provide a threshold of statistical confidence, a variant is reported. In this example a Cytosine is altered to a Thymine at chromosome 5, position 172,660,039 of the hg19 human reference sequence. Note that other mismatches appearing only once or twice among the mapped reads are not reported and could represent errors in sequencing, errors in alignment, or true variants not meeting the statistical threshold for variant calling. C. Annotation and Interpretation: Functional information is added to describe the predicted role this variant might play in changing the function of a protein product of a gene, in this example the gene is NKX2–5 a key transcription factor in heart development which when altered can cause congenital heart disease. The functional annotation utilizes a model of gene based on a variety of information which are organized and maintained by organizations such as the National Center for Biotechnology Information (NCBI) at the National Institutes of Health. In this example the annotation includes the presence of this genetic difference is two importatnt databases of human variation. The ExAC database is maintained by the Broad Institute, and this variant is not present among the 63500 adults who have undergone exome sequencing and are ostensibly free of major malformations and early onset disease. The ClinVar database is maintained by the NCBI and curates genetic variation in genes associated with human disease with information contributed by providers and expert users of genetic testing services; from this resource we see that the variant is present has been reported to be Pathogenic on two separate occasions in two different patients with congenital heart disease. Finally, the qualified clinician involved integrates all available annotations of the genetic variant along to interpret the variant in the context the clinical scenario and determine key next steps; reporting, testing of family members, and the known disease or treatment correlates with the genetic variant.