I’ve been looking at some shotgun data that has been run through MALT (v038) on BLASTN mode to the full NCBI NT Database (from February 2016). MALT parameters: -id 85, -top 1, -supp 0.01, -mq 100.
When loading the RMA6 file into MEGAN6-CE (v6.7.18), and load the alignment viewer on a specific leaf node, not all the alignments that are supposedly on the node are displayed in the viewer. This also occurs on a colleagues version of MEGAN-CE (v6.4.0).
I include a link with an example RMA6 file and screenshots from v6.7.18 here.
To reproduce: If you open the RMA6 file and look for the ‘Ceratotherium simum simum’ node, it is indicated that there are 280 (summarised) hits. If you open the ‘Inspect’ window on that node, you can see the BLAST output of all 280 reads. However if you open the alignment viewer on the same node, there are only available 40 reads to be viewed across three reference sequences.
Is this a bug, or is there some additional filter when using the alignment viewer?
the most likely explanation is that, by default, the alignment viewer only displays a reference sequence if there is at least 10 reads aligned to it. So most of your reads align to reference sequences that attract less than 10 reads.
You can change the threshold to 1, say, by adding the following line
to the MEGAN preferences file Megan.def :
MinReadsInAlignment=1
On linux, find the file here: ~/.Megan.def, on MacOSX here: ~/Libary/Preferences/Megan.def, not sure where Windows puts preference files.
Hi Daniel
I am new to the Diamond-Megan-LR pipeline. I’m currently working on ONT reads from microbial samples - diamond .daa outputs are being analyzed in Megan CE. I am having a similar issue to the above eg There are 3 reads summed/2 reads assigned to a species node, however, only 1 read is aligned to the (7) reference sequences.
I’ve tried the inclusion of MinReadsInAlignment=1 into the Megan.def file, but that hasn’t worked. Can you think of any other potential solution/fix to this?
Let me check that I understand: you have three reads summarized to a species node, of which two are assigned to the species node and one is assigned to a more specific node?
That doesn’t sound unusual to me.
Could you select the species node, then use File->Extract to new Document to create a new document only containing the three reads and then give me access to the document?
Hi Daniel
I guess I am assuming that if a read is assigned to a node I will be able to visualize an alignment associated with it (in show alignment).
Please see link to file with summed reads at species node.