targets of a list of miRNAs

Question

[zizigolu]

hi,

because of lack of an miRNA specific gff file for Aspergillus I used whole mirbase for miRNA de novo identification from sRNA-seq data. now I have a list of differential expressed miRNAs from different organisms. I predicted the targets for example by mirDB. for example target of hsa-miR-4454 was BAG5 a gene in human therefore does it make any sense reporting this genes while the sRNA-seq data come from Aspergillus? if I should homolog of this gene in Aspergillus?

thank you

[Farbod]

Dear Angel, Hi

Have a look at this paper, it compare some miRNA clusters of bat with human's. Of course they are close (in comparison to Aspergillus and human) but maybe some clustering on miRNA families or their target gene enriched GO could help you. By the way, may be the most characteristics of miRNAs is this that they are conserved.

http://bmcgenomics.biomedcentral.com/articles/10.1186/1471-2164-15-682 (look at fig.4 and fig.5)

And maybe this one and this one are good ones to read as they show some bacteria and Aspergillus miRNAs effect on human targets.

Sorry if it did not help.

[zizigolu]

thank you so much for considering my post

[Farbod]

And most of the miRNA target mining databases has skewness to some animals such as human, mouse and rat.

[zizigolu]

sorry,

differential expressed miRNA indentified from Aspergillus sRNA-seq will target Aspergillus genes or human genes?

this sRNA-seq data coming from pathogen itself and pathogen afther human blood infection. then I got confused whether DE miRNAs will target pathogen genes or host (human) genes?

[Farbod]

Hi,

Have you any transcriptomic data of Aspergillus? you can map your miRNAs with your RNA-seq data in this case to find some potential targets (I know that in this case you will lose 80% of your miRNA as they locate in intronic parts but as an observation it has its own value).

About your question, you have some DE mirna results and logically it should tell that some miRNAs are up-regulated in your case condition and some others are up-regulated in your control condition (what is those targets are other story, o of course).

What program did you used for your miRNA DE analysis? if you have used mirdeep2 you should map your reads with the Aspergilus genome, yes?

I remember that I have introduced you with a software that has GUI for mirna analysis, did you use that ?

And I did not get the point clearly about "data coming from pathogen itself and pathogen after human blood infection". Your first condition is from Aspergilus smallRNA-seq and what is the second one ?

[zizigolu]

thank you

mirdeep2 did not worked for me I used Osis GUI by "all genome of mirbase 21v". oasis gave me read counts (although because of lack of a mature and haipin file for this fungi, in mirdeep2 I should use whole miRNAs again) and I performed DEanalysis by edgeR.

if I understood properly they performed sRNA-seq from A.fumigatus before and after human blood infection.

https://i.imgsafe.org/9f1e8e0918.png

https://i.imgsafe.org/9f20875377.png

[Farbod]

Hi

It seems that the link of your images are broken. (Angel did you remove other smallRNAs from your miRNAs?)

Have you tried iMir or miRSeq (miRSeq supports the analysis of up to 105 animal species) and check their DE result with your results?

And if you are using all miRbase mature database, you even no need to use any software and I think you can collect your previously identified miRNAs using BLAST. and then you will search for your novel miRNAs.

[zizigolu]

thank you, imir needs perl :(

[Farbod]

So, what is the problem with poor perl ?

Have you read this paper and the reference No. 20 in it (Transcriptomic analysis of antifungal activity by humidimycin over Aspergillus fumigatus ) ?

[zizigolu]

thank you for assigning time for my problem

long time I am trying tools written in perl but I face pure error

I used aspergillus transcriptome and fasta sequence of miRNAs in rnahybrid tool (https://bibiserv2.cebitec.uni-bielefeld.de/rnahybrid) for identifying targets that gave me aspergillus genes. I have another options, using miranda for human targets or mirDB for all species targets. the same vague; whether these miRNA will targets aspergillus genes or host (human) genes

[Farbod]

Which OS are you using on your server?

[zizigolu]

windows in laptop and remote access to fedora Ibox in max Planck institute

[Farbod]

Hi,

I have used miRNAkey on ubuntu and it was very easy to use with DE profile and some graphs and you can use your fastq or fasta files.

I think you can try it. (unfortunately its website which belongs to Tel Aviv University is restricted for Iran). to see if your DE result is simillar with it. if it was OK you can use its pipeline.

As it uses perl and Java (have a look at its requirement below), If you want, you can send me your two files and I will run it on my server for you. .

.

.

Software Requirements - manual installation

Java Version 6 (download here). Burrows-Wheeler Alignment Tool (download here). Fastx-Toolkit (download here). The following Perl modules:

Math::CDF Spreadsheet::WriteExcel Getopt::Long GD::Graph::bars These modules can be installed by typing the following command in the terminal:

$>cpan Math::CDF Spreadsheet::WriteExcel Getopt::Long GD::Graph::bars Alternatively, they can be downloaded and installed from cpan. (Mac users may encounter problems while installing GD::Graph::bars, if this happens please download the latest libgd and install manually).

System Requirements

miRNAkey can run on Linux/Unix or Mac systems, with 64-bit architecture. miRNAkey has been tested on a variety of Linux systems and on Mac OS X 10.5 and 10.6. 4GB or more of RAM is recommended when using the SEQ-EM option on average-sized Illumina sequence files (10,000,000 or more short reads).

[zizigolu]

thank you I have already tried that and failed :(

are you Iranian?

[Farbod]

Hi,

1- Why? it is a very easy to use and has a simple GUI which you can import your two fastq files at the same time and run DE analysis and you can choose All-mirbase, all-animals, Human and so on for your alignment and you can update your mirbas via it automatically.

http://mybs.u.qiniudn.com/wp-content/uploads/2012/11/miRNAkey.jpg

2- Fortunately, yes.

[zizigolu]

actually I don't know what happen but this my miRNAkey running :( :( :(

miRNAkey starting analysis...

Ca_1.fq - alignment to mature_dna_all.fa

Please wait, this may take some time...

Alignment summary:

Total reads: 0

Mapped reads: 0

Uniquely mapped reads: 0 (0.0% of mapped reads)

Multiply mapped reads: 0 (0.0% of mapped reads)

HsCa-Ca_1.fq - alignment to mature_dna_all.fa

Please wait, this may take some time...

lignment summary:

Total reads: 0

Mapped reads: 0

Uniquely mapped reads: 0 (0.0% of mapped reads)

Multiply mapped reads: 0 (0.0% of mapped reads)

Ca_1.fq_HsCa-Ca_1.fq - Differential-expression analysis

Please wait, this may take some time...

diff_analysis_Ca_1.fq_HsCa-Ca_1.fq.xls Differential Expression Analysis

summary: One or both of the input files have no mapped reads. Quitting....

Note: miRBase database used for this analysis was last updated on Mon Apr 30 12:53:56 2012

Output in: /usr/home/izadi/miRNAkey_out_9_10_2016/

sh: line 1: 14184 Segmentation fault (core dumped) bwa aln -n 2

/usr/data/nfs6/izadi/miRNAkey/DB_mature/all/mature_dna_all.fa

/usr/data/nfs6/izadi/cbbio-miarma/Examples/basic_examples/miRNAs/reads/Ca_1.fq >

alignment_files/Ca_1.fq_aligned.sai 2> /dev/null

fastx_collapser: Invalid quality score value (char ')' ord 41 quality value -23) on line 8

sh: line 1: 15848 Segmentation fault (core dumped) bwa aln -n 2

/usr/data/nfs6/izadi/miRNAkey/DB_mature/all/mature_dna_all.fa /usr/data/nfs6/izadi/cbbio-

miarma/Examples/basic_examples/miRNAs/reads/HsCa-Ca_1.fq > alignment_files/HsCa-Ca_1.fq_aligned.sai 2> /dev/null

fastx_collapser: Invalid quality score value (char '#' ord 35 quality value -29) on line 4

done!

[Farbod]

As it tells, "both of the input files have no mapped reads. Quitting",

maybe you have not installed BWA correctly! (sudo apt-get install bwa)

one way to find if it is OK or not is to perform updating mirbase via miRNAkey, as it somehow check the BWA first.

And did you worked with sample-Data ? If that runs correctly, everything is OK.

sometimes ago I have worked with the small sample data and I did not used the "clip Adaptor" section and then it showed no mapping.

Are your files heavy (e.g GB)?

[zizigolu]

thank you

I tried many options but no success

fortunately I have already taken want I need by Oasis :)

[Farbod]

By the way, one similar issue has occurred to one of my friends when he tried to run some miRNA data from his portable hard-drive in miRNAkey already installed in his laptop.

He received "no mapped reads. Quitting" and when he transferred the data on his laptop . . . ;-)