On 2016 Aug 22, Vincent J Lynch commented:

We have recently become aware of potential contamination in the Wooly Mammoth samples reported in this paper, M25 in particular (for details please see the preprint by Rogers and Slatkin available at https://arxiv.org/abs/1606.06336). While potential contamination of the M25 and M4 genomes with other Wooly Mammoth DNA may render these samples unusable for some analyses, we do not believe that the results or conclusions reported in Lynch et al. are adversely affected. Specifically our analyses focused on fixed, derived substitutions that occurred in the mammoth stem-lineage. The identification of mammoth-specific fixed, derived amino acid substitutions is unlikely to be affected by contamination of the M25 and M4 genomes with aDNA from other mammoths because these changes are by definition fixed in the genome of all mammoths.

We have reanalyzed our data including the additional mammoth samples reported by Palkopoulou et al. (PMC4439331) and replicated all but 105 (91.3%) of the amino acid changes reported in Lynch et al., including the mammoth-specific TRPV3 substitution. Downstream enrichment analyses were also replicated. We are preparing a manuscript describing these reanalyses and will post a link to the preprint when it is available.

On 2016 Mar 16, Vincent J Lynch commented:

Hi Mick,

We may have been able to use the existing data if they were at high enough coverage and available prior to our sequencing of the Asian elephant genomes. But we were likely sequencing the genomes at the same time as Wilkie et al. (2013), thus they just would not have been available to use. We also sequenced each Asian elephant to ~30x whereas Wilkie et al. (2013) and Dastjerdi et al. (2014) sequenced to 5x and 2.5x (as you obviously know). The lower coverage may have made using the existing data less informative.

Vinny

On 2016 Mar 10, Mick Watson commented:

Hi Vinny

Thanks for the reply. I guess my question would be whether you needed to generate new data on Asian elephant genomes, or could have just used existing data

Cheers Mick

On 2015 Aug 05, Vincent J Lynch commented:

Hi Mick,

Although I was aware of this work, we didn't think it was appropriate to cite Dastjerdi et al. (2014) or Wilkie et al. (2013) because we did not use the data reported in these papers or even characterize the Asian elephant genomes we generated. Rather we just used the Asian elephant data we generated to polarize nucleotide changes in order to identify those that were mammoth-specific.

Vinny

On 2015 Jul 09, Mick Watson commented:

Quite surprised this paper didn't recognise previous efforts to sequence Asian elephant genomes:

http://www.gigasciencejournal.com/content/3/1/12

On 2015 Jul 09, Mick Watson commented:

Quite surprised this paper didn't recognise previous efforts to sequence Asian elephant genomes:

http://www.gigasciencejournal.com/content/3/1/12

On 2015 Aug 05, Vincent J Lynch commented:

Hi Mick,

Although I was aware of this work, we didn't think it was appropriate to cite Dastjerdi et al. (2014) or Wilkie et al. (2013) because we did not use the data reported in these papers or even characterize the Asian elephant genomes we generated. Rather we just used the Asian elephant data we generated to polarize nucleotide changes in order to identify those that were mammoth-specific.

Vinny

On 2016 Mar 10, Mick Watson commented:

Hi Vinny

Thanks for the reply. I guess my question would be whether you needed to generate new data on Asian elephant genomes, or could have just used existing data

Cheers Mick

On 2016 Mar 16, Vincent J Lynch commented:

Hi Mick,

We may have been able to use the existing data if they were at high enough coverage and available prior to our sequencing of the Asian elephant genomes. But we were likely sequencing the genomes at the same time as Wilkie et al. (2013), thus they just would not have been available to use. We also sequenced each Asian elephant to ~30x whereas Wilkie et al. (2013) and Dastjerdi et al. (2014) sequenced to 5x and 2.5x (as you obviously know). The lower coverage may have made using the existing data less informative.

Vinny

On 2016 Aug 22, Vincent J Lynch commented:

We have recently become aware of potential contamination in the Wooly Mammoth samples reported in this paper, M25 in particular (for details please see the preprint by Rogers and Slatkin available at https://arxiv.org/abs/1606.06336). While potential contamination of the M25 and M4 genomes with other Wooly Mammoth DNA may render these samples unusable for some analyses, we do not believe that the results or conclusions reported in Lynch et al. are adversely affected. Specifically our analyses focused on fixed, derived substitutions that occurred in the mammoth stem-lineage. The identification of mammoth-specific fixed, derived amino acid substitutions is unlikely to be affected by contamination of the M25 and M4 genomes with aDNA from other mammoths because these changes are by definition fixed in the genome of all mammoths.

We have reanalyzed our data including the additional mammoth samples reported by Palkopoulou et al. (PMC4439331) and replicated all but 105 (91.3%) of the amino acid changes reported in Lynch et al., including the mammoth-specific TRPV3 substitution. Downstream enrichment analyses were also replicated. We are preparing a manuscript describing these reanalyses and will post a link to the preprint when it is available.