• Develop a curatorial ‘shield’, in the form of a Web-hosted software tool - PrediCtoR - which predicts the probability of successful ancient DNA recovery from archive collections in museums (predominantly focussing on fossilised and non-fossilised bone)
• Make the tool scalable and integrate with additional databases to increase the sophistication and robustness of prediction
• Investigate dominant factors influencing optimal DNA recovery from archive bone specimens

JRA1 will develop a novel decision-making tool to help curatorial staff, internal and external Users to quantify the risks associated with destructive analysis of organic specimens held in museums and herbaria.

Developing the basic PrediCtoR system

The basic tool, delivered via a web-based interface, will predict the success of PCR amplification of DNA using a predictive model of DNA decay. PrediCtoR will be a simple-to-use, inexpensive-to-operate, non-destructive tool designed to enhance the conservation of museum and herbaria collections, by preventing unnecessary destructive sampling. It does this by predicting the success of PCR amplification based upon sample size and amplicon length

JRA1 will advance the current state-of-the art by normalising four of the key contributing factors to DNA (and collagen) decay, namely (i) age (ii) temperature (iii) specimen size and (iv) polymer (amplicon) length into a single parameter - thermal age (kyrs@10°C). ‘Thermal age’ normalises different samples and reports them on a common scale; in general the older the specimen the greater the projected accumulated DNA damage. Once these mathematics have been completed it will be encoded in PHP programming language and provided as a simple website accessible for testing principally by Access users.

Proof of Concept

The first part of the proof of concept is to establish that PrediCtoR successfully predicts temperatures for known sites. JRA1 will compare PrediCtoR estimates of T_eff with manually calculated values from actual sample sites and estimates taken directly from weather station data (n = 200). Outliers will be investigated further to establish reasons for any apparent anomalies. Two problems are obvious altitude and changing climate. JRA1 will i) explore the problem of altitude, and incorporate the ICOA standard atmosphere lapse rate (of 6.4 °C km^-1) into PrediCtoR and ii) produce estimates of Pleistocene temperature fluctuations. In order to achieve the latter long-term estimates of Pleistocene air temperature fluctuation (fixed to MAT estimates for the site) will be scaled using estimates of maximum depression at Last Glacial Maximum to estimate T_eff for pre-Holocene samples.

The second part of the proof of concept will be to use large datasets of geographically located samples with age, DNA success and amplicon length, to derive our first real estimate of the rate of decay of DNA in archive bone. This provides PrediCtoR with its first useful output, namely, the first prediction of percentage success (as a function of thermal age).

Task 1.3 Enhancement

The basic core of PrediCtoR will be extended to increase functionally. Initially this will be the implementation of a new collagen prediction algorithm, using an activation energy of 173 kJ mol-1. Collagen is another helical biopolymer, which is also prone to hydrolysis and subsequent melting, but as collagen makes up almost 50% of the total volume of bone, it is much easier to measure using non-invasive methods than DNA. Collagen content prediction can be used to identify exceptionally good/bad preservation of bone indicating predicted success of DNA recovery (see JRA2).

PrediCtoR will normalise thermal age for amplification length and subsequently report to the User both the (a) the maximum amplicon lengths and (b) minimum sample sizes for which 25% 50% and 75% success rates are predicted. This can be computed using standard probabilistic modelling; observed fragment length patterns from non-targeted pyro-sequencing are consistent with random hydrolysis, and DNA concentration, as a general rule, directly scales with sample size. PrediCtoR aims to improve thermal age estimate by taking into account burial depth of the bone (this would invoke an exponential decay function on seasonal fluctuation) and soil type (by including a thermal diffusivity term).

The final version of the PrediCtoR Website will contain the facility to upload results of DNA extraction successes (which will also include details of PCR and ultimately sequencing success as the three elements are all part of the same result process to attain meaningful data). Equally importantly, records of failures will also be invited to enable wider understanding of the challenges and provide a platform to highlight where positive outcomes have been achieved for both User and collections managers. All potential users of museum archive specimens will be able to input information on a given specimen (including geographic location, type, size, extraction protocol, amplification protocols along with projected amplicon length and PCR amplification success.

Work started in Sept 2009. Nothing to report yet.

Work started in Sept 2009. Nothing to report yet.

There are no planned events at this stage.

Each of the PrediCtoR enhancements will be made available as soon as they are developed so the tool can be accessed by researchers.

Individuals may request to upload (current and archive) comparative data on temperatures in their museum stores, which can be used to estimate the effect of storage temperature on long-term DNA viability. Data from other fossil and bone collections will be sought as part of other projects.

For more information contact synthesys@nhm.ac.uk.