On 2018 Jan 17, Ivan Shatsky commented:

The review contains an extensive bibliography and in general could be useful for many readers, especially those who are taking their first steps in studying the molecular mechanisms of translation initiation in eukaryotes. The main and serious drawback of the article is an incomplete and distorted presentation of the current status of cellular IRES-elements. I already wrote in Pub Med Commons critical notes on the article from the same lab (Xue S et al. 2015. Nature 517 (7532):33-38)devoted to the same problem but, unfortunately, I was not heard. This review,again,does not say anything about a severe controversy in this field, it does not even mention the fact that the vast majority, if not all,of the cellular IRESs proposed to date must be validated since they all have been indentified using the method of bicistronic DNA constructs, a method extremely prone to almost unavoidable artifacts.

 Most of the works on individual cellular IRES quoted in the review belong to the period from the late 1990s to the end of the first decade of 2000s. In recent years, however, several articles have been published that point to uncertainties of our knowledge on cellular IRES, describe various pitfalls in the approaches used to identify them and suggest methods to avoid artifacts and obtain reliable information. The main requirement in these approaches is to use various RNA constructs, rather than bicistronic DNA, to transfect cells and assess the level of their expression. These approaches and stringent criteria to analyze  experimental  data were described in several of our papers (see, for instance, Andreev et al. 2009. Nucleic Acids Res. 37: 6135–6147; Shatsky et al. 2010. Mol.Cells 30: 285–293; Andreev et al. 2012. FEBS Lett. 586: 4139–4143; Terenin et al. 2017. Cell Mol Life Sci. 74: 1431–1455) and employed in our lab to verify IRESs in c-Myc, Apaf-1 and LINE-1 mRNAs. The authors of this NRMCB article probably found these papers not worthy of attention since they are not quoted. The only exception is reference 101 which represents the article with a detailed analysis of the status of cellular IRES by 2013. The analysis was performed by Richard Jackson, a widely known expert in eukaryotic translation (Jackson 2013. Cold Spring Harb.Perspect. Biol. 5: a011569–a011569.) Remarkably,although this reference is present in the text, nothing is said about the actual content and conclusions of Jackson's work. 

 The authors note that “over 100 proposed IRES-containing mRNAs have been reported” but “after decades of work, few examples have been well characterized”. However, the authors do not express any surprise why the progress is that slow, why the methods which have been successfully used to validate viral IRESs are not employed or not applicable to characterize cellular IRESs. The authors also note that putative cellular IRESs have “few structural similarities to each other” and that “their mechanisms of action are largely unknown”. Again, the authors do not ask the question:  why for many years now we have not seen positive data on the validation of cellular IRES? Maybe most of these mysterious structures are simply not IRESs, and we must look for alternative mechanisms? Meanwhile, scientific journals, including top ranking ones, go on publishing papers with novel unvalidated cellular IRESs, thereby only thickening the fog in this area 

In short, it seems to me that this section of the review is written in an overly optimistic manner. It is unlikely to be useful for professionals in the translation mechanisms and may be  misleading for new generation of researchers who do not have a sufficient background in the field.

On 2018 Jan 17, Ivan Shatsky commented:

The review contains an extensive bibliography and in general could be useful for many readers, especially those who are taking their first steps in studying the molecular mechanisms of translation initiation in eukaryotes. The main and serious drawback of the article is an incomplete and distorted presentation of the current status of cellular IRES-elements. I already wrote in Pub Med Commons critical notes on the article from the same lab (Xue S et al. 2015. Nature 517 (7532):33-38)devoted to the same problem but, unfortunately, I was not heard. This review,again,does not say anything about a severe controversy in this field, it does not even mention the fact that the vast majority, if not all,of the cellular IRESs proposed to date must be validated since they all have been indentified using the method of bicistronic DNA constructs, a method extremely prone to almost unavoidable artifacts.

 Most of the works on individual cellular IRES quoted in the review belong to the period from the late 1990s to the end of the first decade of 2000s. In recent years, however, several articles have been published that point to uncertainties of our knowledge on cellular IRES, describe various pitfalls in the approaches used to identify them and suggest methods to avoid artifacts and obtain reliable information. The main requirement in these approaches is to use various RNA constructs, rather than bicistronic DNA, to transfect cells and assess the level of their expression. These approaches and stringent criteria to analyze  experimental  data were described in several of our papers (see, for instance, Andreev et al. 2009. Nucleic Acids Res. 37: 6135–6147; Shatsky et al. 2010. Mol.Cells 30: 285–293; Andreev et al. 2012. FEBS Lett. 586: 4139–4143; Terenin et al. 2017. Cell Mol Life Sci. 74: 1431–1455) and employed in our lab to verify IRESs in c-Myc, Apaf-1 and LINE-1 mRNAs. The authors of this NRMCB article probably found these papers not worthy of attention since they are not quoted. The only exception is reference 101 which represents the article with a detailed analysis of the status of cellular IRES by 2013. The analysis was performed by Richard Jackson, a widely known expert in eukaryotic translation (Jackson 2013. Cold Spring Harb.Perspect. Biol. 5: a011569–a011569.) Remarkably,although this reference is present in the text, nothing is said about the actual content and conclusions of Jackson's work. 

 The authors note that “over 100 proposed IRES-containing mRNAs have been reported” but “after decades of work, few examples have been well characterized”. However, the authors do not express any surprise why the progress is that slow, why the methods which have been successfully used to validate viral IRESs are not employed or not applicable to characterize cellular IRESs. The authors also note that putative cellular IRESs have “few structural similarities to each other” and that “their mechanisms of action are largely unknown”. Again, the authors do not ask the question:  why for many years now we have not seen positive data on the validation of cellular IRES? Maybe most of these mysterious structures are simply not IRESs, and we must look for alternative mechanisms? Meanwhile, scientific journals, including top ranking ones, go on publishing papers with novel unvalidated cellular IRESs, thereby only thickening the fog in this area 

In short, it seems to me that this section of the review is written in an overly optimistic manner. It is unlikely to be useful for professionals in the translation mechanisms and may be  misleading for new generation of researchers who do not have a sufficient background in the field.