Cerebellar ataxias are a group of rare neurological disorders that are clinically and genetically heterogeneous, with several hundred genes and diseases known to date. Over the last decade, their diagnosis has been revolutionised by the development of high-throughput sequencing technologies such as exome/genome sequencing (ES/GS), making it possible to obtain a molecular diagnosis in a growing number of patients. However, almost 40% of patients remain without a molecular diagnosis, raising questions about the limitations of sequencing technologies based on a technique known as short-read. One limitation of short-read is its poor ability to detect repeated motif expansions, a frequent mechanism in neurology and associated with more than thirty neurogenetic diseases. Although tools for analysing ES/GS data have gradually been developed in response to this problem, their effectiveness and reliability remain moderate. To date, the gold standard for detecting these expansions remains targeted approaches such as PCR and Southern blot, which are long, tedious and costly processes that require an independent search for each expansion, forcing clinicians to select expansions and limiting diagnostic yield. In addition, there are diseases associated with expansions so rare that no French laboratory offers a diagnostic test.

The recent development of long fragment genome sequencing (long-read - lrGS) could provide a solution to all these problems. These technologies are based on a sequencing process during which DNA is preserved in the form of large molecules of several tens of thousands of bases. Regions of the genome containing expansions can therefore be studied directly in their entirety, avoiding the difficulties of reconstruction from small fragments, which is the case in short-read sequencing. In addition, lrGS can characterize the size of repeated motifs and thus detect any causal expansion in an individual in a single analysis.

A number of recently published studies, particularly in neurology, have demonstrated the ability of lrGS to detect pathologies with known expansions (SCA36, C9ORF72), but also to discover new ones and thus explain the molecular basis of rare pathologies (SCA27b, NOTCH2NLC). Although these sequencing technologies have been around for a number of years, access is still restricted to research work and is limited by their higher cost.

Their value as a second-line diagnostic tool has yet to be demonstrated. The investigators propose to evaluate the feasibility and diagnostic yield of Oxford Nanopore lrGS in duo or trio (patients + 1 or 2 first-degree relatives) in patients with cerebellar ataxia without molecular diagnosis after short-read GS. This will be the first study to transfer this lrGS technique to the second line, in real-life conditions, for the causal genetic diagnosis of cerebellar ataxia.

Inclusion Criteria:

• Index case with progressive cerebellar ataxia of familial form (> 1 1st or 2nd degree relative affected) or sporadic form (onset of symptoms before age 50)
• Index case having undergone srGS and not having obtained a molecular diagnosis, whose srGS data are available for reanalysis.
• Ability to understand and sign consent by the index case and his/her relative(s) (up to a maximum of 2)
• Sample may be taken from the index case and at least one affected or healthy* first-degree relative (parent, sibling) * Healthy relatives must be older than the patient to avoid conducting a presymptomatic test in subjects who consider themselves to be healthy.

Exclusion Criteria:

• Index case or relative(s) not affiliated to national health insurance;
• Index case and his/her parents presenting a condition that, in the opinion of the investigator, would contraindicate the subject's participation in the study.
• Person under legal protection (curatorship, guardianship)
• Person subject to a measure of legal protection
• Pregnant, parturient or breast-feeding women
• An adult who is unable to give consent