The path taken by Todisco et al., as first reported (2025) in this preprint, is very different from the path my laboratory took and has expanded on. Happily, when different researchers, having followed different paths with different organisms, arrive at similar conclusions, one's confidence in the validity of the conclusions increases. Rather than adding a comment here on the preprint, in 2025 I emailed the authors directly.

Quoting them, I noted that they had contributed important information to a fundamental question posed much earlier. Our work had shown that RNA "sequences exhibit clear signatures of selection to counteract" a propensity both to self-aggregate and to avoid internal duplex stem-loops. In addition, they were alerted both to the extensive literature and to a minor reference error. I appended Chapter 3 from my book on Chargaff’s rules (2016) and a recent review in J Theor Biol (2024). This discussed purine-loading and a controversial sequence randomization method we had developed in the 1990s (similar to the one they had employed). I was warmly thanked for my input.

After a 2026 version of their preprint, the paper is now formally published (Proc Nat Acad Sci September 14, 2026). However, on checking I found that the reference had not been corrected (Seffens and Digby 1999, not just Seffens 1999). Kindly added had been a reference to my work (1995). However, the content of my first and subsequent 2025 emails had not influenced the text. Indeed, being familiar with the literature in the field, I noticed another citation error (Workman and Krogh 1999). I did not check for further errors.

In summary, there are errors in substance (references) and scope (coverage of literature on a topic). Had there not been such a long period between preprint and publication, it would be easy to attribute errors to health issues or other distractions. Consistent with this, the authors are unresponsive to my further mailings. However, they have found time for further bioRxiv preprint postings. Thus, I here submit a comment retrospectively. It may be seen as the preprint for a fuller comment to be submitted to Proc Nat Acad Sci – other distractions permitting.

Unannotated Coding Sequences and Novel Transcripts

The finding here of "an unexpectedly broad landscape of tumor immunogenicity" is not unexpected in some quarters. That the peptide aggregates that determine MHC affinity (Scand J Immunol. 2023;98:e13306) are not confined to genic regions is discussed in Chapter 15 of my textbook (Evolutionary Bioinformatics, 3rd edition 2016).

Regarding the puzzlement expressed in this interesting preprint publication on three base periodicity, perhaps Chargaff's rules should be considered? Chapter 6 from the 3rd edition of my textbook Evolutionary Bioinformatics (2016) deals with the vulnerability to cleavage of purine-rich regions. And Chapter 15 deals with the non-genic reality of the so-called "Junk DNA."

Here is an excerpt from Chapter 6 that deals with purine-rich or pyrimidine-rich base clusters:

"It is easy to dismiss such clusters as trivial. After all, in an average audience you may see several members of the same sex together in a row. Males and females are often not evenly distributed. There are statistical tests for telling whether clustering is often not evenly distributed. There are statistical tests for telling whether clustering is likely to be a chance phenomenon. Chargaff found that clustering of pyrimidines was more than expected on a chance basis. Furthermore, with knowledge that DNA was usually in duplex form – pyrimidines in one strand pairing with purines in the other – it was easy to infer that purine clusters must complement pyrimidine clusters."

For information on PubMed Commons see the Laboratory Page (i.e. Lists of Recent Publications, etc).

This page is the main page through which the other pages are accessed, and to which they redirect when finished. Some pages have directions to the PubMed Commons pages that in 2013 began facilitating the annotation of articles in the bioscience literature. Sadly, this was stopped in 2018, but comments can be retrieved through the Hypothesis site as detailed on my Laboratory Page.