Friday, September 12, 2014

Was Maddy's DNA in the hire car?

So we come to the all important question:- Was Maddy's DNA in the hire car?
Perhaps we need to phrase this question a little better before we move on. I prefer:- Was DNA from Maddy present in the sample taken from the boot of the McCann's hire car?

This question has already been answered in the FSS report >John Lowe said the answer was "I don't know."  (I've put that in "" for effect I'm not saying those were his precise words) It is an accurate answer as far as it goes, but as I will show here John Lowe could have said more, MUCH MORE.

Who's DNA could have been in the sample? 

(To be clear, in this post "the sample" always refers to the sample collected from the boot of the hire car)


How many people? 3, 5 or more?

In my previous blog I explained that the sample contained DNA from several people. Lowe says at least three and possibly as many as five. He does not say how he arrives at this conclusion, but I can make a reasonable guess. The "at least three people" almost certainly comes from the observation that some loci contained five (or possibly six) different markers. The only way 5 markers can be found at one locus is if at least three peoples DNA is in the sample.This is because each individual person can contribute a maximum of two markers and minimum of one marker to each locus. Hence 5 markers requires at least three people (2 contributing 2 markers and 1 contributing 1 marker).
It appears that Lowe may have gone on to assume that because he never saw more than 5 markers at a single locus a maximum of 5 people's DNA was present (i.e. if each contributed 1 marker). This is possibly true, in a practical sense, as it is extremely unlikely that DNA from more than 5 people would produce a profile with a maximum of 5 markers at any locus, but in fact it is theoretically possible for the maximum number to be more than 5.

One question we might hope to answer is whether the number of people contributing to the profile was 3, 4, 5 or more than 5. Assuming that the most markers seen at any single locus was 5 I calculate that it is most likely that only three people contributed to the sample, but cannot rule out the possibility that it was more than three. The maths is complicated by several factors. 1. Each marker occurs at a different frequency within the population. 2. We don't know whether the contributors were related or how closely they were related. 3. We don't know how many loci had 5 markers present in the sample. Hence it is not possible to say any more than it was probably three people, but could have been more than three.

Which people have DNA in that sample?

It would be easy to say that the DNA could have belonged to anyone. This is simply not true. The DNA can only have come from a limited number of people who had been in contact with the car or the sample after it was collected from the car. It would have helped greatly if several samples had been taken from different parts of the vehicle & all subjected to LCN analysis. The fact that this was not done is a great pity as it might have allowed a simple subtraction analysis to have been performed. Here is how it works:-

The group would include people at the car hire company, McCann's (including Maddy) + friends & family and forensic scientists + others who had hired the car.

Anyone whose profile has a marker from the 8 loci that is NOT present in the sample can be eliminated. This would probably take care of most people. All of the people remaining would be potential donors to the sample. A potential breakthrough for the case could be made at this point. If it turns out that Kate & Gerry McCann are both potential donors to the sample it becomes highly probable that Maddy's DNA is not present. Their combined DNA would account for all the 15 markers from Maddy's profile. It would not rule out completely that Maddy's DNA was also present, but the odds on this being the case would fall dramatically. Furthermore if it was shown that Kate and Gerry McCann could NOT be contributors to the sample it would greatly increase the chance that Maddy's is in the sample along with all that that implies.

The group of people who are potential contributors could then be further analysed in a group inclusion analysis. Here is how it works:-

All possible combinations of three people are generated. This might be quite a large number of groups, but still possible to do by hand and easier with a computer. All combinations that do not recreate the exact profile seen can be rejected. Most combinations should be rejected by this process that is just a statistical fact. The FSS report is somewhat disingenuous about this when it says that many people are potential contributors to the sample including the report writer Lowe himself. This may be true, but what he does not say is that only a very few combinations of three people who had access to that vehicle or the sample would generate the EXACT profile that was obtained. 

Without access to the full profile obtained from the sample and the DNA profiles of people who are potential contributors to the sample (i.e. those who had access and who's DNA profile fits with the 37 observed markers) it is impossible to know how many groups of three people would be found. It is likely to be a very small number, possibly even zero and almost certainly no more than 5.

If there are no groups that can account for the sample profile it suggests that at least one individual is missing from the analysis or that the sample contained DNA from more than 3 people. The analysis can now be repeated looking at groups of 4 & 5 in fact this should be done regardless of the findings for groups of three.

If there are groups that can account for the sample profile their composition should be studied. There are three possibilities.

1. All the groups contain Maddy.
One key question is do all the groups that account for the sample profile contain Maddy? If the answer is yes then we have once again greatly increased the odds that Maddy's DNA IS present in that sample. Of course this is still some way short of proof, but it would be an indication that the parents should be considered suspects and the possibility that Maddy's body was moved using the hire car.

2. Some groups contain Maddy and some do not.
If there are groups of three people that can product the complete profile of 37 that do not contain Maddy as well as groups that do contain Maddy it would not rule out the possibility that Maddy's DNA was present in the sample, but it would provide a clear and obvious explanation for the presence of the 15 markers without Maddy's body ever having to be in the car thereby shifting suspicion away from the parents.

3. None of the groups contain Maddy.
The final possibility is that only groups not containing Maddy can account for the sample. This is highly unlikely, but if it were to happen it would strongly suggest that Maddy's DNA was not present in the sample.

As I said earlier the group inclusion analysis should be repeated for groups of 4 & 5 people in order to get a clear picture of which groups of people could have produced the observed profile. The composition and number of these groups should then be studied. For example if the only group that can account for the profile is Maddy + a Portuguese  forensic scientist + an FSS forensic scientist this would almost constitute proof that Maddy's DNA was present in the hire car. Alternatively if a group comprising Sean + Amelie + John McCann could account for the sample profile this would provide a strong indication that the sample from the car boot did not contain Maddy's DNA.

Could Maddy's DNA have been in the car without her body being there?

The simple answer to this question is "yes". It is theoretically possible that her DNA might have been transferred from luggage or clothing that she had used or worn to the car boot. However this is rather unlikely. It is also important to note that the DNA sample was obtained from the spot in the car boot to which the CSI dog alerted. So while DNA transfer cannot be ruled out it can be considered unlikely.


1. Analysis of individual DNA profiles and the profile of 37 markers found in the sample can ELIMINATE a large proportion of possible donors and IDENTIFY a smaller group as genuinely potential donors to the sample.

2. If Kate & Gerry McCann are eliminated as potential donors to the sample it greatly increases the probability that Maddy's DNA is present in the sample.

3. If Kate & Gerry McCann are identified as genuinely potential donors to the sample it greatly reduces the probability that Maddy's DNA is in the sample.

4. It should be possible to identify at least one group of 3,4or5 people whose DNA could have been on the car boot sample whose combined DNA profiles match exactly the profile obtained from the sample. The composition of this group or groups might greatly increase or decrease the probability that Maddy's DNA is present in the sample.

5. Most important is the fact that this type of analysis and conclusions is still possible today. The full 37 marker DNA profile from the sample should be available (if it has been destroyed someones head should roll). DNA samples could be obtained from all people who used that car while in the McCann's possession and earlier. Forensic scientist/tech DNA profiles should be on file. 


We are not quite finished with the car yet. My next post will consider what else could have been done in 2007 to confirm or not the presence of Maddy's DNA in the car.


  1. Thanks very much for trying to explain, I have definitely understood more than I normally do.

    It's good to have another expert on this case; we really could do with a lot more.

    So if I've understood correctly, in 5. you are saying that certain tests should have been carried out, in order that a conclusion MIGHT have been reached.

    So doesn't this beg the question why that simple piece of further analysis has not been carried out?

  2. What *kind* of DNA was found in the car?

  3. It was nuclear DNA (from the cell nucleus) found in every cell in the body, i.e. it was not mitochondrial DNA (mtDNA). The difference between the two is well explained in this link: