Novel algorithms for PFGE bacterial typing: Number of co-migrated DNA fragments, linking PFGE to WGS results and computer simulations for evaluation of PulseNet international typing protocols

Abstract

<jats:title>Abstract</jats:title><jats:sec><jats:title>Background</jats:title><jats:p>Standard protocols for Pulsed-field gel electrophoresis (PFGE) were adopted in a global scale for molecular sub-typing and outbreak investigations for many pathogenic bacterial species. Whole genome sequencing (WGS) is playing significant rules for the same mentioned purposes and much more. But the results of the two typing methods are being compared separately. In PFGE; matched bands are considered regardless of co-migration of different DNA fragments. A new algorithm for PFGE image analysis that enabled calculating the number of DNA fragments in each PFGE band is described. The algorithm is based on pixel densities of single DNA fragments. Reliability of the method was assessed <jats:italic>in-silico</jats:italic> using computer simulation approach. 931 whole chromosome sequences (representing 5 bacterial species) were digested by 11 restriction enzymes for each species as recommended by PulseNet international. The simulation was carried out at two levels of accuracy named critical co-migration thresholds (CCTs) assuming that at the first CCT; DNA fragments differ by &gt;= 10% are separated while at the second CCT, fragments differ by &gt;= 5% are separated.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>Total number of bands at 10% CCT was 21,395 PFGE band but at 5% CCT the figure increased to 29,381 bands (by 37.33%). Single bands were found to comprise 64.2% and 45.8% at 5% and 10% CCT respectively. PFGE coverage of &gt;= 95% of whole chromosome sizes comprised 75.26%. While, 83.17% of all digestion models have more than ten and less than 30 bands. A prototype database for comparing PFGE results to whole chromosome enzyme digestion models is deployed (web link: <jats:ext-link xmlns:xlink="http://www.w3.org/1999/xlink" ext-link-type="uri" xlink:href="http://geltowgs.uofk.edu">geltowgs.uofk.edu</jats:ext-link>).</jats:p></jats:sec><jats:sec><jats:title>Conclusions</jats:title><jats:p>The described PFGE image analysis algorithm reveals number of DNA fragments rather than just determine whether a band exists or none exists. Our simulation results have shown reasonable similarities when compared to previously published wet-lab PFGE results. Linking PFGE data to WGS digestion models may further enrich our knowledge about bacterial evolution, population dynamics in addition to reducing the high costs of WGS. Simulation approaches presented in this work may provide a better framework for in-silico evaluation of currently used or future PFGE protocols.</jats:p></jats:sec>

PulseNet international: the missing link between PFGE and WGS

“…Epidemiological conclusions returned by the database were analyzed. Detailed descriptions of their suggested methods are cited (10). The author will also discuss a qualitative approach to analyze GelToWGS results.…”

“…It may seems possible to link PFGE to WGS data. A recently published scientific article (10) identified the following obstacles for doing so;…”

“…Authors of the same article suggested an image analysis algorithm that quantitatively reveal number of co-migrated fragments across PFGE profiles. They named the new parameter “factor of co-migration (FCM)” (10). In order to assess reliability of FCM calculations, they developed “WGSToGel”.…”

“…They deployed geltowgs.uofk.edu; a freely accessible online database. It compares PFGE-FCM image analysis results to in-silico digested whole chromosome sequences (10).…”

“…Xba I digestion of S. enterica serotype Braenderup (strain H9812) suggested by Hunter and her team (10,15).…”