Hi Sachin - The warnings that RDP4 gives you do not mean that the recombination events that the warnings are associated with are not valid or real.  RDP4 is simply warning you that it is unsure of where the breakpoints are located or unsure of which of the sequences is recombinant - it is possible to detect a real recombination event with a high degree of confidence (with multiple different methods and with a low associated p-value for each of the methods) while at the same time the positions of the recombination breakpoints can be very uncertain or the parental sequences can be uncertain or it can be uncertain which of the sequences in an alignment is the actual recombinant.

Thank you sir for your suggestions.

I tried swapping of parents even making event screening more stringent by increasing number of test but still it is showing warning as possible Mis-identification of recombinant and in most of the cases it is showing that one or both of parental sequences as recombinant. So still I have question that, whether such events can be considered as valid recombination events??

Thanks again.....

These are about 26 viral genome sequences each is of about 2.7 kb.

Hello Sachin, I have been trying to use the RDP4 program for a while but I don't seem to know how to use it. Can you suggest any tutorial or video or online help that can help with the basics? Many thanks.

Your can get all information in RDP3 manual or at 

DOI 10.1007/978-1-59745-251-9_9

Good luck

Many thanks, Sachin for your useful advice!

Hi Sachin - The warnings that RDP4 gives you do not mean that the recombination events that the warnings are associated with are not valid or real.  RDP4 is simply warning you that it is unsure of where the breakpoints are located or unsure of which of the sequences is recombinant - it is possible to detect a real recombination event with a high degree of confidence (with multiple different methods and with a low associated p-value for each of the methods) while at the same time the positions of the recombination breakpoints can be very uncertain or the parental sequences can be uncertain or it can be uncertain which of the sequences in an alignment is the actual recombinant.

Hi,  I find  the simplest way of sorting out the many reciprocal relationships in RDP analyses is to make the assumption that the best supported phylogenetic anomaly is probably the most recent recombination.  So, run RDP with the complete set of sequences, and find which sequence has a recombination event that has the largest probability from the largest number of methods - RDP may  find that more than one sequence gives this result with a recombination event in the same (or closely similar) place.  Remove that (or those) sequences.  Now re-run RDP with the smaller number of sequences, and see whether there is (or are) any more sequences that give significant evidence of recombination.  The remove that (or those), and so on.  This procedure of sequential removal starting with the statistically most significant usually also sorts out the reciprocal relationships.

  Another trick is to use only the silent sites, if there are enough.  My colleagues and I have just had a paper accepted for publication in Virus Evolution on potato virus Y, in which we used these strategies.  If you want a prepub - let me know.

the latest version is RDP 4.82 and works quite well in windows 10. the latest manual from the site attached here.

homepage for RDP is given here.

please cite as

Martin DP, Murrell B, Golden M, Khoosal A, & Muhire B (2015) RDP4: Detection and analysis of recombination patterns in virus genomes. Virus Evolution 1: vev003 doi: 10.1093/ve/vev003[PDF]

thanks.

http://web.cbio.uct.ac.za/~darren/rdp.html

Use RDP sequentially as described in Gibbs et al Virus Evolution paper on PVY and its necrogenic recombinants - find the recombinant(s) that is/are most likely (most methods, smallest p). Remove those and repeat test. Unless you do that there is chaos.

The following article will help a lot....

Many thanks for that. I will read it when I get a moment. I'm interested in trying to find non-recombinants so that I can, hopefully, get a 'clean' tree - so I focus on the recombinant sequences, not the recombinant events, but I expect Darren and his colleagues are correct, but I hope we all are. Adrian

Hi, Sachin. I am new here and interested in the question. Actually I've met similar questions. I think you must figure that out after such a long time. Could you please share the final solution with me? Thanks in advance.

Hi, Sachin. I met a problem when I use RDP4. Could you please give me some suggestion? I use RDP 4 to detect recombination events between different samples. And there are 3 recombination evets with major parent unknown. Did you meet the problem before?

If you use RDP4 and rely on the first result, you may get an answer that is not correct, especially if two or more of the recombination events share parents. You can then try sequentially removing the sequences listed in the 'events', starting at the top of the list, and rerunning the program each time. However that does not always solve a problem. The only certain way to check is to remove from the alignment the region that includes the minor parent sequence, and then recalculate and compare trees to see the relationships of the remaining parts.

If the major parent is "unknown" then try BLASTing some bits of it in Genbank as it is obviously not among the sequences in the alignment you are putting into RDP, but may be in the complete Genbank database. Good luck!!!

Adrian Gibbs I installed the RDP4 beta 4.97 but can't find the "X-over" button, instead, I saw "Run" button with a lot of options.

What option is similar with the X-over button?

Hit the option button, then "default" then decide if your sequence is linear or circular, then "save", finally hit "run" and enjoy the screen until the run is complete. Don't forget that it will only find recombinants if both 'parents' are in the alignment. Good luck.