Sian Ellard, Kevin Colclough, Kashyap A. Patel, Andrew T. Hattersley
Submitter: Serena Pezzilli | s.pezzilli@css-mendel.it
Authors: Serena Pezzilli, Vincenzo Trischitta and Sabrina Prudente
Fondazione IRCCS Casa Sollievo della Sofferenza, San Giovanni Rotondo, Italy
Published February 21, 2020
We read with interest the article by Ellard et al., recently published in this Journal (1).
In this Viewpoint, the Authors address how and to what extent the use of bioinformatic algorithms can be very inaccurate to predict the impact of genetic variants in causing autosomal dominant monogenic diabetes. Conversely, the Authors claim the importance of expert diagnostic interpretation according to several key steps, including the guidelines of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP)(2). Addressing this topic is timely and welcome, given the unprecedented ability to identify genetic variants in our day by means of next generation sequencing. In order to highlight the pitfalls of bioinformatic algorithms in predicting variants’ likelihood of causality, the Authors provided several examples referring to publications that would have used such an approach (first 4 references in Ellard et al.) (1). Among these papers, reference 2 refers to our recent publication, which described a Familial form of Diabetes of the Adulthood (FDA) with a likely autosomal dominant inheritance that is often misdiagnosed as type 2 diabetes (3). By study design, FDA pedigrees were characterized by the lack of variants in known monogenic diabetes genes that were likely causal.
It is unfortunate that Ellard et al. (1) have repeatedly misquoted our paper. In detail, in the abstract and on page 14, second paragraph, they stated that “Recently published examples involving monogenic diabetes demonstrate how pathogenicity prediction algorithms can be very inaccurate for predicting which genetic variants are likely causal of dominant monogenic disease (1–4)”. Indeed, we used prediction algorithms only as a prioritization tool in our filtering pipeline aimed at pointing those variants, which deserved further investigation. Conversely, when the likelihood of causality was at issue, we used the ACMG/AMP guidelines (2) as we clearly stated in the last paragraph of page 138 and the first 4 lines of page 139. Once more, on page 14, last two lines of the first column and the first four lines of the second column, the Authors misquoted our paper as one of those over-relying on in silico prediction tools for identifying “likely pathogenic” monogenic diabetes variants (1). As said before, this was definitively not the case.
Reference misquotation, which has long been known (4, 5), may be misleading for the readers and harmful to science. The authors should be very careful both in interpreting and reporting others’ data.
References
1.Ellard S, et al. Prediction algorithms: pitfalls in interpreting genetic variants of autosomal dominant monogenic diabetes. J Clin Invest. 2020;130(1):14-16
2. Richards S, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-424.
3. Pezzilli S, et al. Insights from molecular characterization of adult patients of families with multigenerational diabetes. Diabetes. 2018;67(1):137-145.
4. de Lacey G, Record C, Wade J. How accurate are quotations and references in medical journals? Br Med J (Clin Res Ed). 1985;291(6499):884-886.
5. Jergas H, Baethge C. Quotation accuracy in medical journal articles-a systematic review and meta-analysis. PeerJ. 2015;3:e1364.