I want to find out the naturally occuring plasmid sequences in my bacteria. I have done whole genome sequencing of bacterial genome using Illumina platform.
Did you upload the entire fastq file(s), or the de novo assembly? The first often gives time out errors if your internet connection is not super. If this is the case, try doing a de novo assembly and submitting the assembled genome.
You can also try a more manual approach, by mapping the known plasmids of your organism against the de novo assembly. Usually, plasmid DNA has a higher coverage and different GC content compared to your chromosomal DNA, so this is also a key point that you can look at. So if you get a contig that shows a high similarity to a known plasmid, has a higher coverage and a different GC content than the majority of contigs, you can be quite sure you are looking at a plasmid.
You can also reverse the process, and map your de novo assembly against a reference genome, only containing chromosomal DNA. Then look at the pieces that don't match, if you get a large (or several) contigs that don't map, have a higher coverage and different GC content, you could be looking at a (or several) plasmid (s). You can then blast the sequence(s) to see if you find a hit.
It is not clear to me what exactly you want to do. Do you want to filter out a plasmid you introduced into the bacteria and then continue your analysis with the whole genome data? Or do you want to know which plasmids naturally occur in a bacteria you isolated from a sample?
In case of the second, take a look at PlasmidFinder: https://cge.cbs.dtu.dk//services/PlasmidFinder/
This is a freeware tool that looks up the replicons in a plasmid database on your sequence. The company I work for is also developing a plasmid detection based on the whole plasmid sequences, but this is still under validation and won't be free once we release it.
You could try to assembl using Mira for example and afterwards with Gap5 look after contigs with a very high coverage. Those contigs probably belong to a plasmid and therefore can be identified and removed.
Dear Dr. Vranckx, I want to find out naturally occurring plasmid in my bacterial sample..I tried plamid finder but it didnt work, giving error repeatedly..
Did you upload the entire fastq file(s), or the de novo assembly? The first often gives time out errors if your internet connection is not super. If this is the case, try doing a de novo assembly and submitting the assembled genome.
You can also try a more manual approach, by mapping the known plasmids of your organism against the de novo assembly. Usually, plasmid DNA has a higher coverage and different GC content compared to your chromosomal DNA, so this is also a key point that you can look at. So if you get a contig that shows a high similarity to a known plasmid, has a higher coverage and a different GC content than the majority of contigs, you can be quite sure you are looking at a plasmid.
You can also reverse the process, and map your de novo assembly against a reference genome, only containing chromosomal DNA. Then look at the pieces that don't match, if you get a large (or several) contigs that don't map, have a higher coverage and different GC content, you could be looking at a (or several) plasmid (s). You can then blast the sequence(s) to see if you find a hit.
Previous answers seem exhaustive, I only want to add that in many cases the replication origin of plasmids is known and you can look for that in your assembly.