Genome analysis can make difference

Next generation sequencing (NGS) technologies have revolutionized our capability to analyze the human genome, not only in research projects, but also to aid diagnostics and clinical patient care. Quite often colleagues who are seeing patients with hard-to-define disease ask whether a genome analysis would help in their diagnostic problem.  

In my view genome – or exome, which is the protein coding part of our genome – analysis should be the first tier diagnostic test, if one suspects a genetic disease known to be caused by several different genes. In some cases a more targeted analysis panel is also an option, but particularly, if the genetic background of the disease is not fully characterized (i.e. novel causative gene defects are published frequently), the targeted panels are likely out of date and don’t include all possible candidates.

One patient group benefiting probably the most from a genome analysis is very young children with severe disease. They may present only a small part of the clinical characteristics of a disease, or even have misleading symptoms, which makes selecting the right traditional clinical tests or gene panel for confirming the diagnosis challenging. We and others using NGS in diagnostics have seen that in approximately half of the patients in whom a likely molecular diagnosis was identified by genome analysis, the diagnosis based on traditional clinical tests was refined or altered.

It’s all about interpretation

NGS is today a very reliable and repeatable technology to identify the over 3 million single nucleotide variants we each have in our genome. Copy number alterations, complex structural variants and sequences containing long repeats are still a challenge to current technologies, but further development of algorithms and third generation sequencing platforms likely resolve these challenges in the near future.

However, the interpretation of the genome data is still a major challenge. We all carry 10000-15000 variants potentially altering the protein function. Several bioinformatic tools have been developed to estimate the significance of the observed variants, but the estimates for other than loss-of-function (LoF) variants don’t always agree, and none of the algorithms is significantly better than the others. On average 2-3% of the identified variants may be deleterious for the protein function, translating into hundreds of variants predicted to be deleterious in each person. After these filtering steps, we can say that the 100-300 LoF variants and the approximately 60 de novo variants (variants not seen in the genome of the parents) observed in each individual are the most likely causes of a disease. However, not all of them are harmful, and some LoFs may even be beneficial. So what next?

Value of the population-specific reference data

Reliable estimate of the population frequency of the identified variant helps in deciding whether a variant is a likely disease causing, thus the large sequencing efforts of the population cohorts are extremely valuable also for clinical decision making. It is important to have access and utilize the population specific allele frequencies, such as those provided by the Sequencing Initiative Suomi (SISu). Such data is particularly important for small founder populations like Finns, with both disease-related and neutral alleles enriched in the population.

The actual functional consequence is only known for a very small set of variants, which have been linked to diseases, and have been extensively functionally tested. And even in those cases the Mother Nature can surprise us: depending on the genetic background and different environmental exposures an individual has encountered, the resulting phenotype can be very variable. Even the same mutation can cause two very different phenotypes depending on whether the individual was born with the mutation or whether she or he acquired the mutation in early life. Utilization of genome analysis in research and diagnostics has widened our view on the phenotype spectrum that can result from a defect in a single gene. Every day we learn something new!

Reaching the molecular diagnosis provides several significant clinical advantages: most of all it ends the expensive and time consuming diagnostic odyssey. It also enables more accurate prognosis of the disease, genetic counselling of the family, better estimation of the recurrence risk for next children, family planning, pre- and perinatal diagnostics, screening and early intervention of close relatives who might not yet show the symptoms, and in best cases, early and potentially more efficient and/or targeted therapy.

diagnostics, NGS, sequencing, bioinformatics, rare diseases
Last updated: 31.08.2016 - 13:40


Research Director