We created several shiny R modules for identification of microbial taxa using different molecular markers such as SSR/microsatellites, MLST, and dominant or binary markers (AFLP, RFLP, etc.).
Each of the modules are open source, fully customizable and easy to use, in order to provide the researcher with a modular web-application that permits flexible strain identification for virtually any microbial species.
Quoting the shiny Rstudio tutorial:
Shiny apps have two components:
* a user-interface script: The user-interface (ui) script controls the layout and appearance of your app.
* a server script: The script contains the instructions that your computer needs to build your app.
In our cases, for each of the the components we have a server script (`server.R`) and an HTML based user interphase (`www/index.html`)
The user interphase communicates with the R application by using forms. Each of the forms will contain the query information that will be passed to the R app and be analysed with the reference datasets. Each of the components is fully customizable and contain instructions to customize them efficiently
Genotype-ID is based on R, it uses:
- The latest R version
- [poppr] (https://github.com/poppr/poppr) package and its dependencies.
- [Shiny server] (https://github.com/rstudio/shiny-server) and its dependencies
After installing the requirements, be sure that you have a folder named shiny-server on /var/. Download the files into /var/shiny-server/.
After downloading the files, start the shiny-server
$ sudo shiny-server &
And go to your preferred browser (Genotype-ID is known to have issues with Internet Explorer, so we recommend using Google Chrome or, better yet, Firefox) and go to the server port address:
http://localhost:3838/
The browser will show you a link to a folder called "Genotype-ID/", if you click on it, the application will deploy on shiny server and voila! there is your web-app.
Genotype-ID is a strain identification tool based on SSR. Genotype-ID uses poppr to construct a Bruvo's distance dendrogram (using UPGMA/NJ) or a Minimum spanning network of an SSR query against a well-curated database of reference SSR's.
popprpackage
Tab-delimited text file. Column names are loci names, rownames are samples. Each of the alleles is separated by a /. The second column contains the reference population/lineage/code to group reference strains into discrete groups.
Here's an example of the reduced_database.txt.csv file that is used by Genotype-ID as reference:
| Sample | Lineage | PrMS6 | PrMS9c3 | ILVOPrMS132/PrMS39a | PRMS45 | ILVOPrMS133/PrMS43ab | KI18 | KI64 | ILVOPrMS131 |
|---|---|---|---|---|---|---|---|---|---|
| PR-12-044 | NA1 | 165/168 | 216/226 | 216/226 | 372/489 | 130/246 | 219/275 | 342/379 | 150/234 |
| PR-11-015 | NA1 | 165/168 | 216/226 | 216/226 | 372/481 | 130/250 | 219/275 | 342/379 | 154/222 |
| PR-11-008 | NA1 | 165/168 | 216/226 | 216/226 | 368/485 | 130/250 | 219/275 | 342/379 | 150/226 |
At the user interphase, a text form will be filled containing the same loci for different queries:
sample1 ??? 169/172 221/231 136/252 172/192 373/489 224/278 347/383 150/234
sample2 ??? 169/172 221/221 152/158 0/0 173/173 226/226 349/365 278/342
- Customizable minimum spanning network
- UPGMA/NJ dendrogram
MLST-ID is a MultiLocus Sequence Type Id app created in shiny R. It uses the information of multiple FASTA files of reference organisms to provide clear and fast identification of microbial species using MLST.
R packages:
apephangornXMLphyloch
Additional software
The MLST-ID uses a FASTA reference per gene/locus of interest:
Gene_A.fasta:
>0310_WAS
CGCGGCAGTCGCCGTCGCCGAGGCGCCGGCCAAGGCATACAACCCTCTCTTCATCTACGGCGACTCGGGCCTCGGCAAGACCCACCTGCTGCACGCCATC
>0202_ECU
CGCGGCAGTCGCCGTCGCCGAGGCGCCGGCCAAGGCATACAACCCTCTCTTCATCTACGGCGACTCGGGCCTCGGCAAGACCCACCTGCTGCACGCCATC
The input is a text form that requires a FASTA input. The database which the query will be compared needs to be specified in the FASTA identifier (>Gene_A):
>Gene_A
CGCGGCAGTCGCCGTCGCCGAGGCGCCGGCCAAGGCATACAACCCTCTCTTCATCTACGGCGACTCGGGCCTCGGCAAGACCCACCTGCTGCACGCCATC
The query will be aligned with the database of the reference sequences for the gene of interest using MAFFT, and then the selected genetic distance will be calculated to construct the dendogram/minimum spanning network of the query/database.
- Customizable minimum spanning network
- UPGMA/NJ dendrogram
Binary-ID uses codominant molecular markers (coded as 1 and 0) and different genetic distances to reconstruct a distance dendogram with bootstrap support values and a minimum spanning network.
popprR package
A dataset of AFLP loci of Aphanomyces eumithes of two different populations, as included in poppr was used a an example dataset. Aeut.txt: Column names are each of the AFLP site profiles. Row names are sample names. Second column are the populations of the reference samples:
| Ind | Pop | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
|---|---|---|---|---|---|---|---|---|
| 1 | Athena | 1 | 1 | 1 | 0 | 0 | 0 | 0 |
| 2 | Athena | 0 | 1 | 1 | 0 | 0 | 0 | 0 |
| 3 | Mt. Vernon | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
| 4 | Mt. Vernon | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
| 5 | Mt. Vernon | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
At the user interphase, a text form will be filled containing the same loci for different queries:
Ind_1 Queue 1 0 1 0 1 0 0
Ind_2 Queue 1 0 1 0 0 0 0
- Customizable minimum spanning network
- UPGMA/NJ dendrogram