top of page

The Merle Gene

Merle Collies are genetically sable or tricolour, with the Merle modifying gene overlaid. You can read more about coat colour inheritance in the Collie here. We know that the Merle gene cannot affect pheomelanin, the red or brown pigment seen on sable Collies, therefore sable merles most often look phenotypically (to our eye) sable. Merle cannot be carried, it is a dominant gene (although incompletely dominant). This means that a dog is either Merle or they aren't - it cannot lay hidden and be produced down the line in offspring, from non-Merle parents. However...

The Merle gene is not as simple as 'non-merle' and 'merle'. For many years, breeders have known that there was more to Merle than meets the eye. The breed standard describes the desired Blue Merle colour: 'silvery blue', which is 'splashed and marbled' with black patches. 'Large black markings' are undesirable, so is a darker, 'slate colour'. But, the perfect silver-blue merle can sometimes produce offspring with large patches, a dark slate colour, or dark undercoat. What causes this? Recent research has unveiled the complexities of the Merle gene, and provided us with a simple genetic test to identify various types of merle.

"Blue Merle: predominantly clear, silvery blue, splashed and marbled with black. Rich tan markings preferred, but absence should not be penalised. Large black markings, slate colour, or rusty tinge either of top or undercoat are highly undesirable."

- Kennel Club

Merle has been mapped to the SILV gene, a pigment gene. A SINE insertion has been found on the gene. To put it simply, this means each Merle gene has extra DNA tacked on the end of it. This extra DNA, called a poly-A tail, can vary in length, and it is the length that determines the features of that particular animal's Merle pattern. 4 distinct groups of Merle poly-A tail length have been identified. We will look at these groups and their effects later in this article.

Merle Gene Testing

Most laboratories simply test for the presence of Merle, they are not able to identify how long the poly-A tail is. One of the only labs to test poly-A tail length, and the only recommended lab, is Tilia. Tilia are recommended as they use the full range of 6 different Merle variants.

To order a test, select 'Molecular Genetics' from the 'List of Investigations / On-Line Orders' tab of the website. Click 'On-Line Orders - Molecular Genetics', and find the 'Merle Coat pattern SILV SINE insertion (m/Mc/Mc+/Ma/Ma+/M/Mh) test. At the top of the page, you will need to fill in some information about your animal, your address and owner information, plus the type of sample you will be taking. We will discuss the various sample types later in this article. 

When you have completed this form, you can add the test to your basket and checkout. You may pay Tilia for their services via PayPal, a quick and convenient option. However, there are various other ways to pay the laboratory, and you can check their website or contact them for more information.

When you have ordered your test, you will need to collect your sample, and send it to the laboratory via the address you were provided after ordering.

Merle Test Results

When your sample has been analysed, you will receive either via email or post, a certificate containing your Collie's test results. Below is an example of what you will receive. 


As previously mentioned, the Merle poly-A tail lengths are grouped into 6 distinct groups. For the test results above, a poly-A tail length (size of the Merle allele) of 267 base pairs was identified.

The chart below shows the different base pairs (test results) and the group of Merle they fit into. So, for example a test result of 273 base pairs would be an Mh result. A test result of 230 base pairs would be an Mc result.

Merle Sine Insertions.png

Merle Types

The features of the 6 different Merle types are described below. The Merle types are defined on a continuum, but it is important to note that the higher or lower base pair values can start to 'act' like the Merle type it is close to. For example, a base pair of 254 can act very similar to Ma+, due to how close to it it lies on the continuum - but is still classified as Ma. The key is the fluidity of the continuum.


No Merle pattern is expressed. Without testing, we would think these Collies were sable or tricolour without the Merle gene. This is described as 'Cryptic Merle'.

When homozygous (Mc / Mc), no Merle pattern is expressed, however the dog may appear to be slightly off colour.

Merle Sine Insertions.png


No Merle pattern is expressed. Without testing, we would again, think these Collies were sable or tricolour without the Merle gene. This is called 'Cryptic Merle +'

When homozygous (Mc+ / Mc+), like with homozygous Mc, there is no change to the coat colour, but the dog may appear slightly off colour. For result Mc / Mc+, the same affect is achieved - no Merle patterning is expressed, but the colour might be slightly off. 


No Merle pattern is expressed. They may sometimes appear to have a diluted coat, similar to the colours achieved by the Dilute genes b and d, which are not present in the breed. There is a Collie colour known as 'Maltese Merle', with some of these dogs being tested as Ma - 'Atypical Merle'. Maltese Merles look like tricolours but with a grey colour where they would usually be black. This covers all of the black, and there are no black patches in the coat.

Dogs who are Atypical Merle can have blue eyes or blue-flecked eyes (heterochromia).


Homozygous Atypical Merle (Ma / Ma), or Atypical Merle with a second Cryptic Merle gene, are often mostly diluted with very small dark patches. For example a Blue Merle might be mostly grey with small black patches.


The Merle pattern can be visible, but not desirable. Coat is often a darker grey, not clear blue, with black patches not being crisp. Sometimes Ma+ dogs have no visible merle. 

Dogs who are Atypical Merle+ can have blue eyes or blue-flecked eyes (heterochromia).


Homozygous Atypical Merle+ (Ma+ / Ma+), or Atypical Merle+ with a second Cryptic Merle+ gene, may express a tweed pattern, particularly as the base pairs move closer to the M group. Tweed pattern occurs when a Collie may have an extra colour in it's Merle coat, for example a light grey, black, and an extra darker black patch. Pigment can be deleted to white and blue or blue-flecked eyes can express.

Ma+ / Mc can present similarly to m/Ma+. Similarly, sometimes no Merle pattern is expressed. No pigment is deleted to white and blue or blue-flecked eyes can express.


The Merle pattern that one thinks of as typically Merle - some pigment is lightened and the coat has black patches. A desirable silvery blue colour can be achieved with M. Pigment is not deleted to white although the closer the base pairs to the Mh zone, might potentially add the chance of pigment deletion, causing white patches. Blue, or blue-flecked eyes can occur.


Homozygous M (M / M) can cause the 'Double Merle' effect, with the dog appearing almost solid white, sometimes with a few patches of black or grey, and with occular and/or auditory issues.

M with a Cryptic Merle can appear as 'classic Merle', with the latter pigment can be deleted to white. M with an Atypical Merle often presents with an extension of Irish Spotting. Pigment can be deleted to white. 

Juliet m M.jpg


m/M (267bp)

Tested by Hair Sample

Sent in by Atlanta Jackson


Mh, or Harlequin Merle, Herding Harlequin (not to be confused with the Harlequin / H gene in Great Danes) can present in different ways. Sometimes, the Collie will have large black patches and only small areas of diluted pigment. Irish Spotting can be extended, such as with a large white collar. Other effects that can occur with Mh include larger white patches, sometimes with a Blue Merle Collie appearing white with black patches only.

Sometimes, these dogs appear as the classic Merle. This may include smaller white patches or extension of the Irish Spotting. When bred, a Harlequin Merle Collie's offspring will not always appear the same 'type' of Mh. 

M / Mh and Mh / Mh often present with the 'Double Merle' effect - with the dog appearing almost solid white, sometimes with a few patches of black or grey, and sometimes with occular and/or auditory issues.

Mh in combination with another Merle gene lower than M in the continuum can express in different ways, sometimes with white patches or large white areas, tweed patterning, etc.



Windkist’s Lean On Me, CGCA, CGCU, SPOT-ON, RATI, RATO, TKP

m/Mh (273bp)

Tested by Hair Sample

Sent in by Shawn Ashley Evans

Mosaic Merle

Another thing to consider is Mosaicism. Sometimes, a mutation occurs where the poly-A tail in a cell shortens, this occurs during the early stages of embryonic development. This cell replicates, and thus this mutation can be found in certain areas of the dog's body. It will not occur in every cell, as the mutation occurred in an already replicated cell - and there are still cells with the original poly-A tail length left. This means a dog can have 2, 3, or potentially even more, different Merle gene lengths. This is not a rare event and occurs in a relatively large proportion of Merle dogs.

The shortened allele is usually a 'minor allele'. In test results, it is shown enclosed in square brackets, for example: 'm/Mh/[M]'. However, sometimes the mutated, shortened allele replicates faster than the original allele. In this case, the original Allele will become the 'minor allele'.

Since the mutation which shortens some poly-A tails occurs early in development, it is sometimes possible for these cells to be found in Germline Cells (located in a bitch's eggs or a dog's sperm). When a shortened allele is found in these cells, they can be passed on to offspring.

The Merle SINE insertion paper identified that out of all the dog's tested, 18% were mosaic. They identified that the longer the poly-A tail, the more likely it was to mutate.

Octavian mos m Mcplus 244 Mh 277 Mcplusm


mosaic m/[Mc+(244bp)]/Mh(277bp) and S/S

Tested by Buccal Swab

Sent in by Jessica C. Gerard

Merle Types

Samples for Testing

Different samples can be taken and sent to the lab for Merle SINE insertion testing.

Shortened poly-A tails (mosaicism) can be found in certain areas of the body and are not found throughout all cells. Therefore, some samples might not show mosaic results when in reality, the dog might be mosaic.

For example, in the Merle SINE insertion study, a hair sample was taken with a m/Mh result. A swab sample resulted in a m/[Mc]/Mh genotype. Another example occurred when a blood sample displayed no mosaicism, whereas a hair sample returned a mosaic minor Mc+ allele. 

Therefore sometimes a sample will return a result not expected when viewed in relation to offspring results or the Collie's phenotype. Sometimes taking another, different, sample, will display the expected mosaic result.

If a male dog has produced offspring with a Merle phenotype or genotype that doesn't seem possible for his tested result, a sperm sample is recommended by the Merle researchers. This will show the Merle results present in his Germline cells, and therefore what he is capable of passing on.

Other sample options are buccal swab, blood and hair. Removed dewclaws could also be used. For bitches, sperm samples are not an option, so there is no way to test the Germline cells. However, testing of her offspring could reveal any undetected mosaic results.

Samples for Testing

Merle and Dermatomyositis

In research conducted by Dr Leigh Ann Clark et al. three risk alleles for the prevalent autoimmune skin and muscle condition DMS in the Collie were found. They discovered that every Collie tested was at risk of DMS, and that statement still holds true even with far more Collies having been tested since.

DMS risk is calculated by a string of six letters. A capital letter is a risk increasing allele, and a lower case letter is an allele that does not convey additional DMS risk. The different combinations have different effects, and you can read more about this here


An interesting fact to note is that an 'at risk allele' for DMS, which is called A in the study, was not found in conjunction with the merle coat colour gene: 'Merle was only found in conjunction with a [the not at risk allele]. Accordingly, the Merle phenotype was underrepresented in affected dogs'.


In practice this means that merle dogs will always have at least one a (not 'at risk'). The researchers determine that selection for a merle coat pattern would be beneficial to reduce the incidences of DMS. In other words - merle dogs have a decreased risk of DMS.


This doesn't mean that we should only breed merle collies. But knowing this fact about the merle gene and DMS is yet another tool in our belt when it comes to tackle the issue of DMS in the breed. It also provides another case for the breeding of merle dogs to sable dogs. Merle can only affect eumelanin (black pigment), which is barely displayed in pure homozygous sables, meaning that sable merles are often phenotypically sable (they look like usual sable dogs to the eye).

Merle and Dermatomyositis

This article is copyrighted by iCollie and must not be reproduced without prior consent. Please share the link, found at the top of your browser, to share this article.

Further Reading

The Merle SINE insertion research is well worth a read yourself, particularly if you are a breeder. There are 3 formats through which you can learn more.

The most technical is the research paper, found here

You can read some of the information on the Merle - SINE Insertion from Mc - Mh website.

The most convenient way to read the studies is within the Merle book, written by Mary Langevin, who you will find wrote the research paper. You can order the book here. It is available via eBook (PDF Document) or hard copy; the hard copy is printed in sets and you can pre-order for the next print. If you do this, you will also be sent the eBook via email. To read the eBook, you will need a PDF reader such as Adobe Acrobat. You can download Adobe Acrobat for free here. (Correct link at time of writing). 

More information on Merle and DMS can be found here.

bottom of page