Who creates genome sequencing applications

Genetic Diagnostics: The "$ 1,000 Genome"

Illustration: Jane Ades / NHGRI

New DNA sequencing technologies are finding their way into medical practice and are increasingly changing patient care.

Soon after the first near-complete sequence of the human genome was published in 2001, the term “$ 1,000 genome” was making the rounds. Behind this was the hope of significantly lower costs for DNA sequencing. At the same time, it was expected that comprehensive sequencing of the human genome could quickly find its way into routine medical care.

Sequencing is available for relatively little money

The vision from back then is now on the way to becoming reality. The sequencing of the two 3.2 billion base pairs of a human genome in a quality sufficient for medical purposes currently only costs around 10,000 euros. From around 2,000 euros, all 180,000 human exons can be sequenced, i.e. those medically relevant sequence areas of the approximately 23,000 protein-coding genes that together make up around 1.5 percent of the genome. And for less than 1,000 euros, the exons of genes that are responsible for known monogenic diseases can be sequenced. The prerequisite for all these examinations in the clinical context is the existence of a corresponding (human genetic) indication - for example, the suspicion of a monogenic disease, the presence of a developmental disorder or an indication of the presence of a syndrome.

The technological quantum leap in DNA sequencing is due to the development of laboratory procedures, which are all based on the principle of parallelization of the enzymatic reactions required for sequencing in the smallest of volumes. Together they are referred to as “Next Generation Sequencing” (NGS). Compared to classic Sanger sequencing, NGS enables costs to be reduced by two to four orders of magnitude, i.e. by a factor of 100 to 10,000! This drop in prices set in motion an enormous methodological upheaval in human genetics.

In patient care, however, an NGS-based genome analysis requires not only the purely technical generation of sequence data but also a competent bioinformatic analysis and the final medical diagnosis. The first two steps can already be largely automated today. The third and last step, namely the care-appropriate interpretation of the detected gene variants, on the other hand, has become the most time-consuming and cost-intensive part of medical genome analysis. It absolutely requires the expertise of a human geneticist or another appropriately qualified specialist.

Lots of data - but little clinical knowledge

After the genomes of a large number of individuals have been completely sequenced in recent years, it is becoming more and more apparent that causal genetic changes can only be read directly from the DNA sequence in a small number of future patients. On the other hand, the clinical significance of the variants mainly revealed in one patient will initially be unclear, as their connection with the clinical picture in question (“genotype-phenotype relationship”) has not been adequately scientifically proven. A comprehensive medical diagnosis of NGS data will therefore not be possible in the near future due to the lack of evidence alone.

In order to remedy this situation in the long term, extensive and publicly accessible databases are required in which disease-causing gene variants are systematically recorded and stored with the associated medical data and the respective evidence of their relevance to the disease. Genotype-phenotype databases of this kind cannot be created simply by using existing registers or disease-specific mutation databases - on the one hand, because individual gene variants can show cross-disease effects, on the other hand, because existing registers do not meet the requirements of a practical, either in terms of content or IT usable NGS database.

Standards for the classification of the relevance of gene variants were developed years ago by human geneticists. As a rule, they are based on the penetrance of a mutation, i.e. on the certainty with which this mutation predicts a certain disease. On the one hand, estimates of the penetrance of a mutation can be obtained statistically by examining a large number of carriers of the mutation. On the other hand, penetrances can also be estimated experimentally using cellular or in vitro models. This has long been practiced, for example, in the classification of variants of the HIV protease with regard to their sensitivity to pharmacological protease inhibitors.

However, there is one problem: Most of the currently available human mutation databases were primarily set up for scientific purposes and are only conditionally suitable for use in patient care. In particular, they lack the detailed clinical background information necessary for genetic diagnosis and counseling of future patients. Scientifically motivated catalogs of disease-causing gene variants were already created in the era of Sanger sequencing (e.g. the Human Gene Mutation Database). At that time, this was possible with comparably little bioinformatic effort. However, the number of newly identified mutations through the NGS will increase so much in the future that a more comprehensive approach with contemporary methods of direct data transmission and the (at least partially) automated maintenance and quality control of genetic and clinical data will be required.

Systematic cataloging of human gene variants

A worldwide network of bioinformaticians, systems biologists and human geneticists is currently devoted to the systematic cataloging of human gene variants. In addition to (primarily scientifically oriented) projects in Great Britain, the Netherlands and the USA, there is also a German initiative coordinated by the Technology and Methods Platform for Networked Medical Research in Germany (TMF eV) in Berlin. The genotype-phenotype database planned as part of this initiative is intended to systematically store benign and disease-causing genetic variants and their accompanying data in compliance with German data protection regulations and make them available for the assessment of DNA changes found in the future. It is intended as a resource for science and patient care alike.

In a long-term perspective, the EU Commission and the US National Institutes of Health, together with other partners, have started the “International Rare Disease Research Consortium” in order to clarify all monogenic diseases with regard to their underlying molecular defects by 2020 (www.irdirc .org). If this hope is actually fulfilled, a decisive step in genotype-based personalized medicine into the reality of care would be taken.

Just as important as these bioinformatics efforts, however, is the sounding out of the ethical, legal and social framework conditions for the use of NGS technologies in medical practice. The aim must be to protect the personal rights of those affected regardless of technological progress and to dispel existing fears of a "geneticization" of medicine in patients, their relatives, the medical professions and in public. The qualification of a large number of doctors for genetic counseling and examinations in accordance with the Genetic Diagnostics Act, currently operated by the state medical associations, makes a significant contribution to this at national level. This ensures that the extensive need for such services to be expected from a possible introduction of the NGS into routine care is adequately taken into account.

In order to support the medical diagnosis of NGS data in terms of content and also to make it cost-efficient for the community of solidarity, trend-setting decisions are required at the moment with regard to the creation of a German genotype-phenotype database and its international networking. The legal framework for such a database in Germany is largely defined by the Genetic Diagnostics Act and the data protection legislation. For this reason, human genetic, epidemiological and bioinformatic efforts will be required in the coming years in order to put the use of NGS as an important part of personalized medicine in Germany on a sustainable basis.

PD Dr. med. Arne Pfeufer

Helmholtz Centre Munich

Prof. Dr. Michael Krawczak

University of Kiel and TMF - technology and method platform for
networked medical research e.V.