Transcript-mapping file

[Chandra Shekhar Misra]

Hi 

First of all thank you for this excellent tool for Alternative splicing determination. Lot of people with little or no computational background may find it very interesting including me. 

I have a question: 

1. In step 1 of Input data of 3D analysis, in tab transcript mapping file, are we suppose to use gtf annotation or .fa file. I tried to upload .fa file from AtRtd2 but it showed me error. Can you please tell me which file to upload in that tab. 

Thanks 

Chandra Shekhar

 

[wenbin guo]

Hi Chandra,

Many thanks for using our 3D RNA-seq App. 

I have tested both .fa and .gtf files of AtRTD2 downloaded from the website: https://ics.hutton.ac.uk/atRTD/ with my dataset. Both formats worked fine. Could you please give me more details about what errors you got?

If the input .fa or .gtf file is not consisent to your transcriptome used for transcript quantification, you will get errors for inconsistency. For example, your transcript quantification is based on TAIR10, but you use AtRTD2 for 3D Analysis.

Another solution is instead of using .fa or .gtf file, you can generate a transcript-gene mapping csv file from the transcript quantification results, with first column of transcript names and second column of gene names. In Arabidopsis, the gene names are just the first 9 letters of the transcripts, e.g. transcript is AT1G01060.1, then the gene is AT1G01060; transcript is AT1G01010.1 then gene is AT1G01010.  You can use excel to open the quantification output of a sample (the quant.sf file from Salmon or abundance.tsv file from Kallisto). You will see a transcript name column in these files. Then in excel you can take the first 9 letters of each transcripts to generate a column of genes. Save the transcript-gene mapping as csv file, which can be used as input for 3D Analysis. Hope this can solve your problem. Please see Figure 1C for details in the user manual: https://github.com/wyguo/ThreeDRNAseq/blob/master/vignettes/user_manuals/3D_RNA-seq_App_manual.md#input-files 

Cheers,

Wenbin

[Chandra Shekhar Misra]

Hi Wenbin

Thank you very much the detailed reply. I did use the AtRTD2 from the link you suggested. I will take the screen shot and send you when i redo the analysis. But i liked the another alternative which you suggested and  would like to give it a try. But based on what you suggested, this gene mapping file obtained from quant.sf file generated from Salmon should be kind of universal and can be therefore used for all subsequent analysis. 

Thanks 

Chandra Shekhar

[Chandra Shekhar Misra]

Hi  Wenbin

 I asked about transcript mapping file but looks like it works now. Although I tried another alternative suggestion you gave and it works too.

While i try to generate heat map in "Functional Plot" tab, the server gets hanged. I tried several time but it always the same. I have attached the screen shot. 

Also, the test data that is provided with the App is based on comparing splicing during different cold treatments which means there are several conditions, can we use the app only in pairwise mode?

In 3D analysis- "Make plots of multiple genes", but I don’t find a way to download the gene list. Also, do we only have top 100 3D targets?

Also in step 5 of 3D analysis, what is “Slice profile plot on group” .

 Also, I was curious if there is any way to find out the Proportion of the major types of AS events using this 3D App.

 Sorry for asking these many questions.

 

Thank you

Chandra Shekhar

 

 

--
You received this message because you are subscribed to a topic in the Google Groups "3D RNA-seq App user group" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/3d-rna-seq-app-user-group/tyFwqiXkMHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to 3d-rna-seq-app-use...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/3d-rna-seq-app-user-group/9de36a92-be2e-4622-8e6b-8b92c5848586%40googlegroups.com.

[wenbin guo]

Hi Chandra,

Many thanks for your questions and these questions do help us to make improvement of our 3D App

1. For the transcript mapping, it might be caused by some transcript mismatches in your transcript-gene mapping csv file and the transcript quantification. I have made an update this morning to take the consensus part of these two inputs for the analysis.

2. For the heatmap, I can’t tell what’s going wrong just by the screenshot. I have tried different datasets, the heatmap plot works fine to me. There are two ways we can solve the issue. (1) After 3D analysis, you just skip the heatmap step and go to the final page generate report. Then click first button save all 3D analysis results and click the last button to download results. In your download, you will get an intermediate_data.RData in the “Data” folder. If you can send this object to me, I can check what’s wrong with your dataset. (2) If the data is unpublished, once you get the grey page, could you please go to the data generation page, step 1: Input data of 3D analysis. At the bottom, you will see the directory of a temporary folder. If you can provide this temporary folder path to me, I can check the log information of your analysis. 

3. The App can be used for pair-wise mode, e.g. you only have two conditions.

4. To make plots of multiple genes, currently, there are two places you can download the list (1) In the functional plot GO annotation panel, you can download the 3D gene lists. (2) In the final step, you can click save all 3D analysis  results and then click download results, all the intermediate results will be saved and downloaded to your local computer. There is a “result” folder in your download. It includes all csv files of the gene/transcript lists, the testing statistics of all genes/transcripts (both significant & non-significant), where you can get results with different p-value ranges. In the next version, I will add a function to let users to download gene lists in the multiple plot panel.

5. Slice profile plot on group. If your experiment is performed in different blocks, and each block has several conditions, then you can slice the expression profile plot by the block. E.g. in our dataset

No slice:

Slice by day block

6. The 3D App doesn’t have the function to generate AS events. What you can do is to download the DAS gene and DTU transcript list, then used a third party tool to generate AS events, e.g. Suppa. We have a similar study in our Arabidopsis time-series experiment. See Figure 7 and section “Identification of Isoform Switches” in this publication: http://www.plantcell.org/content/30/7/1424#F8

Hope these answer your questions.  

Best,

Wenbin

To unsubscribe from this group and all its topics, send an email to 3d-rna-seq-app-user-group+unsub...@googlegroups.com.