 Hi everyone, let's start with the practical part of our tutorial. At first we are going to provide a name tour of each story, for example Arabidopsis transcriptant analysis. Then we are going to upload our datasets from Sonodo. In order to do that we are going to use our TAPRA data provided in the training. We should copy it, then select the uploader tool by using the rule basic method, select collection and paste our table of data and build. Now we are going to provide some rules. We should click on the rules, add a modifier rules, add definition, then list of identifiers and select the column A. Now add definition, collection name and select column B. After that select URL and column C and finally add definition, type and select the column D. Now apply and finally upload. Now meanwhile we are going to upload the rates of our datasets. Now we should copy the tabular data of the additional datasets, select the uploader tool, select datasets, paste the tabular data and build. Once again we are going to create two rules, add definition, name and select the column A and URL and select the column B, then apply and upload. We have all all required datasets available in Galaxy. Once the process of uploading has been finished, now we can check the contents of one of the collection. We can open a FATQ file, we can see four lines for each reads, the sequence identification, the reads and the coding, the line and coding the quality. Now we can add a few tags in order to identify your sequence. Now for example in this case we can add brassine asteroids and microRNA. Now everything is ready. Here we can see the additional datasets that we are going to require, which include the annotation, the whole transcriptome and a few additional datasets. The next step is to evaluate the quality of our reads for which we are going to use the tool FastQC. We should open a collection, select the control microRNA, leave the rest of the parameters at default and repeat the same process with the brassine asteroid treated dataset. FastQC provides a simple way to do some quality control on our reads. Now we can check the results of this tool. It generates two different outputs. Here we can see the different information which it provides like the per type sequence quality, per sequence quality scores, per base sequence content, per sequence GC content, per base unknown base content, sequence length, sequence duplication and over-represented sequence. Here we can see some of the adapters which are present in our reads. Now in order to compare the results of the two treatments we are going to merge the outputs of FastQC. In order to do it we are going to use the tool MergeCollection in single list of datasets. We are going to look for this tool, here it is, this one. We are going to select the collection, the control and the brassine asteroid route collections and execute. Now we have the raw reports Merge in a single list. After that we are going to use the MultiQC tool in order to generate a report. Select collection at the Merge one. Now we are going to provide our title, for example, initial quality assignment and execute. In that way we will be able to visualize all datasets in a single report. As we can see, MultiQC generates two types of outputs, an HDLM and a stats file. You can check the stats file, it provides raw information, but the most interesting one is the HTML file. Here we can see the information that it provides. As we can see our reads include a lot of duplicate reads, for example. Also it provides information about the quality scores, the per sequence quality scores. Here we can compare the value of all of datasets. As we can see all of them show similar trends. As we can see a high percentage of all reads include adapters, which will be necessary to remove as they can interfere with subsequent analysis. To remove the adapters we will use the Tringalore tool. Now select Tringalore, collection, control, collection. Select the Illumina small RNA adapters and execute. Now we should repeat the process with the Brazilosteroid treated collection. Select the Illumina small RNA adapters and execute. The next step is to evaluate once again the quality of our reads after the processing by Tringalore. Now we should select the fast QC tool, select the processor reads, control, collection and execute. And we should do the same with the Brazilosteroid treated collection. As we did before we need to merge the output generated by fast QC. Now we should wait a little bit and select merge collection tool. Select the collection of the control of the Brazilosteroid treated collection and execute. Once we get the Merged collection we should once again use the Multi QC tool in order to generate the report. Select Multi QC tool, Merged collection. Now we are going to provide a report title for example post-processing quality assessment and execute. Once we have obtained the quality report of the processing sequence we are going to compare the quality parameters before and after having processed the samples. For which we are going to use the Scratch Book option. We should select Scratch Book and open the visualization of both reports side by side. In that way it will be quite easy to compare them. Here we have the initial quality assessment and the post-processing quality assessment. As we can see we have a quite similar amount of duplicated reads before and after the processing. The quality histograms are also quite similar. The per sequence quality scores have been improved after the trimming. We can also compare the per base sequence content. As we can see the per sequence easy content has been improved in our reads. In the case of the sequence leg we can see that the original reads have all of them similar length. And after the trimming we have reads with different size. The duplication levels are quite similar. Also the over-representing sequence shows a similar pattern and the adapter content. You can see we have removed all the adapters. Now we can start with the quantification of the micro-ironase. It will be done by using the mirror-deep tool. Then let's go to the tool search bar. Here we have three models. We are going to use mirror-deep to mapper and the mirror-deep to quantifier. First we use mirror-deep to mapper. We should select control-processed collection and we should select collapse and execute. Now we are going to repeat the process with the brassinosteroid treated collection. Here it is. We should select it. Once again select collapse and execute. The collapse tool ensures that each sequence only occurs once. To indicate how many times reads the sequence represents a suffix is added to each FASTA identifier. Now we are going to use the mirror-deep quantifier tool. We are going to select first the control-collection. Now we should select the precursor sequence, the mature micro-ironase sequence and the star sequence. All those files are required in order to perform the quantification. Now we should left the rest of the parameters at default and execute. Now we should repeat the process with the brassinosteroid treated collection. Once again we should select each of the datasets and execute. This tool maps the reads to predefined micro-irona precursors and determines by that the expression of the corresponding micro-ironase. Firstly, the predefined major micro-irona sequence are mapped to the predefined precursors. Then the predefined star sequence are mapped to the precursors too. This tool generates two types of outputs. The raw reads, which include the information such as the micro-irona names, the reads counts, the precursors and the normalized counts. And it also provides an HTML file with additional information such as the mirror-based precursors, the mature read counts, the star read counts and the major sequence, the mirror-based star sequence and the precursor sequence. Before performing the differential expression analysis, it is necessary to extract the first two columns from the quantification files. Those columns include information about the micro-irona identifiers and the reads counts. Then we should select the tool called columns from a table. Select the first two columns, select the collection, the control sequence and execute. And we should repeat the process with the brassinosteroid treated collection and execute. Everything is ready for performing the differential expression analysis. We can have a look to the outputs. We are going to use the sec2 tool. We should specify a factor name, for example, the effect of brassinosteroids. Specify the first factor level, which will be the brassinosteroids. Select the collection corresponding to the brassinosteroids, which is the 104. Then specify the second factor, which is the control and select the corresponding collection. Now we should un-set the files have a header option and leave the rest of the parameters after-fall and execute. The sec2 is a popular tool for gene-level differential expression analysis. It uses the negative Venovel distribution, employing a slightly more stringier approach compared to other methods. It provides a good balance between sensitivity and specificity. The sec2 generates two different types of outputs. Plots and the differential expression data file. Let's have a look to each of them. One of the plots generated by the sec2 is the principal component analysis plot, which provides insight into the association between the samples. As we can see, the X-axis includes the 46% of the variances of the samples. Other plots generated by the sec2 are the sample-to-sample distance, which provides similar information to the PCA plot, the dispersion estimates, Instagram of p-values, and the MA plot. Now let's check the tabular data generated by the sec2. It provides diverse statistics, such as the min expression, the fault chain, the standard error, the world statistics, the p-value, and the p-adjusted value. The p-value is a measure of the probability that an absurd difference occurs just by random chance. On the other hand, the p-adjusted value takes into account the fault discovery rate, which is necessary when we are measuring thousands of variables. Now we are going to filter those genes that show statistically significant expression differences between the two experimental conditions. Now we should select the output of the sec2, select the column number 7, which corresponds to the p-adjusted value, and execute. To perform differential expression analysis, it's recommended to use at least three biological replicates of each experimental condition. In addition, another important factor in determining the statistical power is the sample size. Let's take a look at the results. Unfortunately, we haven't identified anything with significant differential expression. This is caused by the fact that our sample size is not large enough, so we have used a two-sample obtained from the original data. Let's repeat this last step by using the full microRNA dataset. We should copy the cenotaphile address, provide a proper name, for example, microRNA, the sec2, complete dataset, and start. Now we are going to repeat the filtering process with this new dataset. Select the previous result, now we are going to select the complete dataset and execute. Let's check the results. When using the original data, the statistical power is significantly increasing, so in the samples are 100 times larger in size. In this case, we have identified 39 genes whose differential expression is statistically significant. Finally, we'll filter out the upregulated microRNAs. Let's select the previously filtered dataset and set the column 3 higher than 0. In this way, we'll keep only those transcripts whose fault chain is greater than 0. Finally, we have obtained 16 microRNAs whose expression shows a significant increase in Plastic Nosteroid Fritted samples with respect to the control samples. And that's all. I hope you enjoy.