Currently, the methods evaluated in (Hücker et al., 2021) start with micro-manipulation to separate single-cells and perform library preparation for the detection of miRNA. Unlike mRNA, miRNA have no Poly-A tails and therefore conventional library preparation methods of single-cell miRNA preparation are not applicable. There are efforts underway to optimize the library preparation methods by manipulating different technical conditions such as buffer concentrations, enzymes used for ligation and the adapters.

For now the Sandberg Protocol with CleanTag adapters appear to give the best results in terms of miRNA capture and reproducibility across different cell culture types (Hücker et al., 2021). However, we offer the option to process alternative adapter configurations as listed below.

The conditions that are observed at the sequencing level, and are therefore relevant to the preprocessing of single-cell miRNA data are the UMI_length, 3’ and 5’ adapters along with the method. * Index_number is also included as a parameter, however this is the conventional Illumina TruSeq index primer and is not expected to play a major part in the performance of single cell miRNA sequencing. The protocols primarily used here (and denoted by Hücker et al., 2021) are Small-Seq protocol denoted SB and SBN (Faridani et al., 2016, Hagemann-Jensen et al., 2018), CleanTag protocol, denoted as CL (Shore et al., 2016), 4N protocol, denoted as 4N (Giraldez et al., 2018) and CATS, denoted as CATS (Turchinovich et al., 2014). A graphical representation of the structure of the reads can be found in Figure 1. Since it is possible to swap the 3’ and 5’ adapters between the different methods, we allow these methods in the corresponding columns.

Settings

Index number (Index number):

Corresponds to TruSeq adapters 1 to 12. Currently, library preparation for Illumina Sequencing as per Hücker et al., 2021 is supported. The index number has to be specified for each file in the metadata sheet as an extra column which has to be provided by the user.

Protocol variant (protocol_variant):

For each sample a protocol variant is needed. The protocol variants can be the same for all files, therefore, we included the global parameter option. The user can also specify individual protocol variants for each file in the metadata sheet as an extra column. Parameters are separately explained under the table.

Table 1: Protocols and their variants supported by the preprocessing (Hücker, S. et al., 2021).

Protocol variant	3' adapter value	5' adapter value	Method	UMI length
SB (Sandberg protocol I)	Sandberg	Sandberg	SB	8
SB_4N	4N	4N	SB	8
SB_CL	CleanTag	CleanTag	SB	8
SB_C3	Sandberg	C3	SB	6
SBN (Sandberg protocol II)	Sandberg	Sandberg	SBN	8
SBN_4N	4N	4N	SBN	8
SBN_CL	CleanTag	CleanTag	SBN	8
CL (CleanTag Protocol)	CleanTag	CleanTag	CleanTag	8
CL_16C	CleanTag	CleanTag	CleanTag	8
CL_4N	4N	4N	CleanTag	8
CL_Block	CleanTag	Block	CleanTag	6
CL_C3	CleanTag	C3	CleanTag	6
CL_Rand	Rand	CleanTag	CleanTag	6
CL_SB	Sandberg	Sandberg	CleanTag	8
CL_UMI6	CleanTag	CleanTag UMI6	CleanTag	8
4N	4n	4N	4N	0
4N_C3	4N	C3	4N	0
4N_CL	CleanTag	CleanTag	4N	0
CATS	CATS	CATS	CATS	0

Details

UMI length

Length of the UMI used for the library preparation method. For example, for the Sandberg protocol (Faridani et al., 2016, Hagemann-Jensen et al., 2018), the UMI was 8 nt long, while for the Sandberg 3’ CleanTag variant, the UMI was shortened to 6 nt.

5' Adapter

• “Sandberg”: Adapters are following the methods of (Faridani et al., 2016, Hagemann-Jensen et al., 2018, denoted SB and SBN respectively. However, since the primary difference between the methods is the modification of technical variables other than the adapters, for the preprocessing, they are treated as one lable.

• “CleanTag”: Modified adapters to inhibit formation of adapter dimers while allowing reverse transcriptase read-through (Shores et al., 2016).

• “4N”: Adapter has 4 random nuclei at the ligation site, as per Giraldez et al., 2018).

• “C3”: The 5’ adapter is complementary fo the 3’ adapter

• “CleanTag UMI6”: The UMI of the 5′ adapter was shortened from 8 nt to 6 nt.

• “Block”: 5’ CleanTag adapter replaced by an adapter with 3 additional uridine nucleotides at the 3’ end and an UMI shortened to 6 nt. The RT primer is altered, able to bind to adapter dimers.

• “CATS”: Adapters corresponding to CATS small RNA sequencing kit for Illumina from Diogenode.

3' Adapter

• “CleanTag”: Modified adapters to inhibit formation of adapter dimers while allowing reverse transcriptase read-through (Shores et al., 2016).

• “Sandberg”: Adapters are following the methods of (Faridani et al., 2016, Hagemann-Jensen et al., 2018), denoted SB and SBN respectively. However, since the primary difference between the methods is the modification of technical variables other than the adapters, for the preprocessing, they are treated as one label.

• “4N”: Adapter has 4 random nuclei at the ligation site, as per Giraldez et al., 2018.

• “Rand”: CleanTag adapters with first 6 nucleotides exchanged with random nucleotides.

• “CATS”: Adapters corresponding to CATS small RNA sequencing kit for Illumina from Diogenode

Method

• “Jensen” : Corresponds to Hagemann-Jensen et al., 2018.

• “Giraldez” : Corresponds to (Giraldez et al., 2018), primarily used in combination with 5’ adapter option C3.

References:

Faridani, O.R., et al (2016) Single-cell sequencing of the small-RNA transcriptome Nature Biotechnology, 34(12), pp. 1264-1266, DOI: 10.1038/nbt.3701

Giraldez, M.D., et al (2018) Comprehensive multi-center assessment of small RNA-seq methods for quantitative MIRNA profiling Nature Biotechnology, 36(8), pp. 746-757, DOI: 10.1038/nbt.4183

Hagemann-Jensen, M., et al (2018a) Small-seq for single-cell small-RNA sequencing Nature Protocols, 13(10), pp. 2407-2424, DOI: 10.1038/s41596-018-0049-y

Hagemann-Jensen, M., et al (2018b) Small-seq for single-cell small-RNA sequencing Nature Protocols, 13(10), pp. 2407-2424, DOI: 10.1038/s41596-018-0049-y

Hücker, S.M., et al (2021) Single-cell microrna sequencing method comparison and application to cell lines and circulating lung tumor cells Nature Communications, 12(1), pp. 1264-1266, DOI: 10.1038/s41467-021-24611-w

Page 1: Data information

In the first step of the uploading, a name for the project should be provided.

Page 2: Data upload

Step 1: Upload fastq data files

• Drop fastq files or click for upload (accepted file formats: .fastq.gz).
• Press "Start Upload" only when all files have been selected and are displayed in the table.
• If two files with the same name are selected, only the second file is used.

   

Step 2: Pressing the next button

• If there was an error with one or more files during upload, all other files with a successful upload will be used for the next steps.
• Press the next button when all files have been uploaded.

Page 3: Metadata upload

Step 3: Specify a metadata sheet

• A template for the metadata sheet (containing the previously uploaded fastq file names) can be downloaded via the button
• One row per data file (column name "file_name" should contain all uploaded fastq file names)
• Required column: "Index number"
• Column "protocol_variant" can be included and adjusted per row. If the protocol variant is the same for all rows, you can specify a global setting in the next step.
• Possible setting values for "protocol_variant" can be seen in the drop down menu of Step 4 (only enter the name up to the colon to the "protocol_variant" column) and on the help page (Table 1) which also helps if some settings are unclear.
• Optional column "ID" for descriptive sample names
• It is possible to add further analysis after the alignment like DE analysis, etc if the number of uploaded fastq files in the previous step exceed one. To do this, additional categories such as age, case/control status, etc. must be added to the metadata sheet as a separate column

Table 1: Metadata sheet of the example data

ID	file_name	Index number	protocol_variant	Condition	cell_line
SRR12535752	SRR12535752.fastq.gz	1	SB	adenocarcinoma	MCF7
SRR12535878	SRR12535878.fastq.gz	2	SB	adenocarcinoma	MCF7
SRR14284493	SRR14284493.fastq.gz	4	SBN_CL	healthy_fibroblast	BJ
SRR14284494	SRR14284494.fastq.gz	5	SBN_CL	healthy_fibroblast	BJ

• Drop metadata sheet or click for upload (accepted file formats: .csv and .tsv).
• Press "Start Upload" when the correct file appears in the table under the upload area.

   

Step 4: Global setttings

• Specify a global value for "protocol_variant" using a drop-down menu
• Chosen value overrides the specification in the metadata sheet
• Possibility to keep the uploaded metadata sheet small and simple

Step 5: Pressing the next / submit button

• Pressing the next button automatically forwards to the settings for further analysis such as DE analysis, embedding, sample correlation and clustering.
• To skip the analysis, check the box above the next button and press the submit button to start the calculations directly and not include any extra analysis.
• Note that it is not possible to add further analysis later if you skip the next step. The results only contain the expression matrix and multiQC report.

Page 4: Selection of categories for further analysis

Step 6: Select the column names to use as categories

• Categories are required for further analysis. Please note that only those categories that have more than one option in the metadata sheet are useful for the analysis and only those are shown here.
• Select the categories you wish to analyse further by marking the checkbox.
• Please also check the correctness of the displayed and comma-separated options that are available in a category. It will no longer be possible to change these after starting the run. For the seeting method we internally use the name "Jensen" for method "SB" and "SBN" and "Giraldez" for "4N".
• If no further category is displayed except the one for the calculations needed ("protocol_variant", "umi_length", "3'adapter", "5'adapter", "method"), no corresponding information was found in the metadata sheet.
  • If a further analysis of the data is wished, go back to the previous step and upload a new metadata file containing additional sample specific information.
  • If no further analyses is wished, go back to the previous step and tick the box at the bottom of the page and press submit to start the run without a downstream analysis.

Step 7 (optional): Specify colours for the category options

• Colours are provided for each category option for visualisation in the analysis.
• By pressing the colour preview, the colours can be adjusted manually by the user.

Step 8: Pressing the submit button

Pressing the submit button at the bottom of the page starts the run.

Result page

Enter job ID

• Enter the job ID which was provided after the job submission to check the status of the run.
• Or press the button which inserts the example data job ID.
• If the run is finshed or if the example data job ID was inserted, the results are displayed.

Downloading and viewing the results

• No downstream analysis:
• On the top of the page is a button to download the alignment results (including the expression matrices and the multiQC).
• Afterwards the multiQC of the data is displayed.

• Downstream analysis:
• On the top of the page is a button to download the alignment results (including the expression matrices and the multiQC).
• And a button to download the results of the downstream analysis (including png and csv files).
• The multiQC is displayed underneath.
• Using the tabs above the results for expression clustering, sample correlation and embedding can be accessed.
• The last tab leads to the DE analysis for which a separat tutorial exists.

Step 1: Select category

Select the category for which the DE analysis should be calculated.

Step 2: Select case options

Select the option within the selected category to specify the case option of the comparison.

Step 3: Select control options

Select the option within the selected category to specify the control option of the comparison.

Step 4: Select adjustment method for pvalues

Select the adjustment method for the pvalues of the statistical tests (t-test and Wilcoxon rank-sum test).

If all specifications are set start the calcualtions with the corresponding button.

DE analysis results

• On the top of the DE analysis section is a button to download the DE analysis results (including a matrix with all calculated values and png files displaying volcano plots).
• Volcano plots are displayed underneath.
• A new DE analysis for another comparison can again be started like explained above.