The authors conducted a genetic investigation of pigmentation in two color morphs of the common brushtail opossum. A GWAS clearly indicated that variation at the ASIP gene is reponsible for the difference between grey (~agouti) and black animals. Further experiments then convincingly demonstrated that a missense variant in the coding sequence, p.Arg115Cys, leads to a recessive loss-of-function allele that causes the black phenotype. The authors further investigate other marsupial genome sequences and find interesting variation at the ASIP gene that might explain e.g. the black pigmentation of the Tasmanian devil.This is the first report of molecular variation at the ASIP locus in marsupials and demonstrates homologies in the regulation of pigment type-switching between eutherians and marsupials. The manuscript gives a lot of background information (e.g. on the significance of opossums for humans) that is not strictly necessary to convey the key findings. The entire manuscript is well balanced, but could be shortened, if there is a need to restrict the length. The findings are of interest to a large audience and I have only minor comments.

(1)
The authors should consider revising their variant nomenclature. In human genetics, the terms "polymorphism" (incl SNP) and "mutation" are no longer used. These ambiguous terms have been repalced by the term variant. The authors are encouraged to strictly adhere to the human nomenclature rules, as this allows a much more concise descriptons of their findings. (The findings could also be described much shorter, as the lengthy verbose descriptions of sequence changes can be simply replaced by appored variant designations. The HGVS nomenclature rules can be found here: https://hgvs-nomenclature.org/stable/

(2)
While the desicription of the methods is largely fine, it would nonetheless help, if the version of the reference genome assembly or accession number could be repeatedly stated, when they are essential (e.g. in tables with genomic positions).

(3)
Lines 60/61: Gain of function of ASIP leads to obesity in only one specific mouse mutant and rare (or possibly a single?) human patient. This should not be generalized. There are several well-known pheomelanistic mutants due to gain-of-function at ASIP, which are not obese (e.g. dominant yellow dogs, arctic wolves, white Merino sheep, Saanen goats, ...). To the best of my knowledge, the vast majority of red-haired humans are due to loss-of-function at MC1R. Perhaps adding the word "rare" should be inserted before human, when you speak about ASIP-related red hairs and obesity.
(From a molecular point of view: If the overexpressed ASIP can also interact with MC4R, you may expect obesity. Genomic alteration leading to a restricted overexpression in the skin result in pheomelanistic pigmentation without obesity.)

(4)
Line 83: I suggest to rephrase: ... that naturally carries --> that naturally occurs in

(5)
Line 138: 0.5 cm^2 of tissue is not very clear. Please specify in milligrams how much tissue you used (or use something verbose like lentil-sized)

(6)
Line 234: amplicons --> amplicon

(7)
Lines 318 ff.: I suggest to use arabic numbering and not roman numbering for the ASIP exons. Pease note that eutherian mammals have multiple 5'-untranslated exons. The exact number varies between species and is often not known. In eutherian mammals, the three coding exons are often termed exons 2-4, while the 5'-untranlated exons are referred to as exon 1a, exon 1b etc. However, there is no real consensus how these exons should be numbered.

(8)
Line 407/408 & figure 4A: The font size in the multiple species alignment of figure 4 is too small. I m unable to read this. In this figure, it should be indicated, which of the cysteine residues is mutated in black mutants of which species.

(9)
Line 448: In what aspect are the two detected ASIP transcripts novel? I suggest to delete the word "novel".

(10)
Line 450: Fine-mapping --> Further analysis

(11)
Line 458: An allele or a trait can be recessive. A variant (or a mutation) cannot be reccessive.

(12)
Lines 539/540: I suggest to rephrase: In summary, we show that a missense variant in the coding sequence and not a quantitative change in expression of the ASIP gene is ...

(13)
Line 834, legend to Figure 1: Delete "(significant SNPs)". The Manhattan plot shows the p-values of all markers and just the significant ones. You should briefly decribed in your legend what the red and blue horizontal lines mean and your rationale for setting these specific significance thresholds. Why do you need two different significance thresholds? I suggest to delete the lower hoizontal line and the less stringent significance threshold.

(14)
Figure 3:
I suggest to add a photo of a red Dunedin opossum as an additional supplementary figure. In the legend of this supplementary figure, please very briefly dicuss your hypothesis about the moelcular mechanism that leads to this red phenotype.