AN_RNASeq

Compare gene expression profiles obtained under different conditions Study genetic profiles from transcript to pathway level Introduction RNA sequencing, the analysis of gene expression profiles by next generation sequencing (NGS), has become a pow - erful tool to investigate the transcrip - tome of an organism. The comparison of two or more conditions (e.g. mutant vs. wild type) allows the identification of differentially expressed genes that are up- or downregulated under a specific condition. Typical applications include the comparison of gene expression pro - files between normal and cancer tissue, cells in high and low nutrient environ - ments, stressed and unstressed cells or cells from distinct developmental stages. Major prerequisites for any RNA sequencing study are the availability of an annotated reference genome or tran - scriptome and sufficient numbers of bio - logical replicates. mRNA Sequencing for Differential Gene Expression Analysis Microsynth’s Competences and Services Differential gene expression analysis by RNA sequencing is one of Microsynth’s core competences. Based on years of experience, we provide a one-stop service from experimental design to bioinformatics analysis (see Figure 1 ). You can either outsource the whole process or only single steps to us. To ensure a high reliability and accuracy of our product, the whole RNA sequenc - ing pipeline including the bioinformat - ics analysis was evaluated and validated based on RNA spike-ins designed by the External RNA Control Consortium (ERCC). For further information on our validation process and possibilities to validate your own study, please contact us. Experimental Design The gain and impact of a study highly depends on its experimental design. The use of controls as well as appropri - ate sampling and RNA isolation methods are only a few examples of points to con - sider. What is important for any RNA sequencing project is the number of biological replicates. To obtain statisti - cally valid results for a differential gene expression analysis, we usually advise to include at least three biological rep - licates per condition. Make use of our experience – our NGS specialists are happy to assist you from the start. RNA Isolation You can either perform the extraction yourself or outsource this critical step to us. Microsynth has extensive experience in RNA isolation from various demand - ing tissues and matrices. Figure 1. Microsynth's workflow for RNA sequencing projects. The workflow can be entered and exited at various steps dependent on the customer's requirements. Microsynth AG, SwitzerlandSch?tzenstrasse 15 ? P.O. Box ? CH - 9436 Balgach ? Phone + 41fi71fi722fi83 33 ? Fax + 41fi71fi722fi87 58 ? info @microsynth.ch ? www.microsynth.com THE SWISS DNA COMPANY Application Note ? Next Generation Sequencing Example Results The major outcomes of the differen - tial gene expression analysis are the observed fold changes for each gene together with statistical measures and the normalized counts per sample (see Figure 2A ). Statistics for all genes are additionally summarized in an interac - tive table together with boxplots visu - alizing the observed fold changes (see Figure 2B ). The interactive table allows you to sort the data by any represented measure or to search for specific features of interest. Library Preparation and Sequencing Following a quality check of the RNA, we will either perform poly(A) enrich - ment or ribosomal RNA (rRNA) depletion, depending on the organism or the target RNA to be studied. This step is essential since total RNA includes a large fraction of rRNA and other non-mRNA species and sequencing should be restricted to mRNA to avoid losses in sequencing depth. A stranded Illumina cDNA library is created by reverse transcription including the ligation of sequencing adaptors with barcodes. Finally, the libraries are pooled and sequenced on the Illumina NextSeq platform with a typical single-end read length of 75 bp. The sequencing depth per sample highly depends on the studied organisms and the desired sen - sitivity. Benchmarks for complex eukar - yotic genomes (e.g. human, rat, mouse) are 100-150 million reads for high sensi - tivity and 20-30 million reads for typical sensitivity. Prokaryotic genomes require approximately 5-fold less reads. Bioinformatics Analysis After quality control, the sequenc - ing reads are mapped to the reference genome using the software Salmon [1] or STAR [2], which both address the diffi - culty of mapping spliced reads. As input for statistical analysis the reads that uniquely map to a gene are counted. The identification of differentially expressed genes is performed by specialized statis - tical software such as DESeq2 [3]. Read counts are normalized, and the vari - ance based on the replicates per con - dition is calculated. Finally statistical testing is applied to identify differentially expressed genes that are significantly up- or downregulated. For organisms with available pathway information, a complementary service is provided identifying significantly up- or downregulated pathways (Gene ontol - ogy and KEGG classification). Figure 2. Summary tables resulting from the dif - ferential gene expression analysis. 2A. Excerpt of a table summarizing the results of the analysis for two conditions with three replicates each. ID: gene ID; baseMean: average number of read counts; log2FC: log2 transformed fold change between conditions; lfcSE: standard error of log fold change; p-value: Wald test p-value; padj: p-value adjusted for multi - ple testing. 2B. The statistical results are also avail - able as an interactive html-table allowing sorting and searching for specific features. 2U6pV
$I?Vr”AHV@c3gG7( 47br@21 &UV6’"p@C3 9t€uYuB%1T @21&UV6’"p@C3 "Qpf3Aap )v)8 uBeY(X 1$a yy•CS•5)„b0$5wv@(Q˜ 76ed6!)v3 2`4g‡ #B2% 9YcG
4‰VB Sx‰H 85S "3pˆ 3b2i 8`”„P'’ˆ&t3V3"`8qY5e BC$$ GH cBGh0x—‘ F9` 4UD%rTt –G8s 7r64 ’’€4„y5u Rb™?G@u?B5gG‘)8p!U5A”gg !aA6 rƒ€(7ƒWP cS•@y2i‚ •04 %“„S”&@ 09 09 #IH) "A70
1@p"06s& 0‘cb 7“–G™ i˜bbr ˜‚h€ Ea9 2
–C76c‚ (†“q †“r PS% i“q'@#v#`UdH1 ‚(W ƒC( )8 ?@8€P7`T $ay„ B‰u„ ?„Bb€–ywB Pp5d 1ƒ$P "vv@ `F“?FI

W‘3?‰%8rsE! B #1D?5„?…8G‰ip9B H !9%f 9W“Fe#BQ€s &TQG P(T 7G —3S†“@ '”0dT?f0`8—tXQ˜…€ Bah‘† tR—… 4i1UA $?26 r™? I—PbgeI2P y…p $S`uP "PTR 7—9˜XY(2SAtAs’IRIe (hu‚3 Pb‘w1YPQ"1u W—q #y$%`&cf" $’i&$ &˜6C dX“ 8C#”9’T‚P&Cˆy79™™Q $%€$82—' `„“P 'c Q6`q? 5hP` 8`‰Tp‘& eE‰p %C™8t‡Ib@"14—T u%
??TeY(X A 8Q€gc# %ˆT%ˆS0 s?V & 0„xuB 'rX9uB #pS" 9†q4 uB 'rX9uB #pS" 9†q4 uB 'rX9uB )eYy –
w9@ tDb 9I”X’ Yt™ 1u4P` B Microsynth AG, SwitzerlandSch?tzenstrasse 15 ? P.O. Box ? CH - 9436 Balgach ? Phone + 41fi71fi722fi83 33 ? Fax + 41fi71fi722fi87 58 ? info @microsynth.ch ? www.microsynth.com THE SWISS DNA COMPANY Application Note ? Next Generation Sequencing UYQ% A„ Y˜‡ƒ0 (”fI $GF3i 7PQ(F !`@t˜ % A„ 0VPT€ 5Fa™‡ uBuBuB 5ES• —f„g` •R B9d?V 6YiWp X(aC Xxht? 3•‘V@ 1x 2b ` A•IH #f—% sˆ8v For each comparison in the differen - tial gene expression analysis, various graphs are provided for an easy and visual overview of the results (see Figure 3). The graphs include MA plots visualiz - ing the distribution of the differentially expressed genes, heatmaps for sample to sample distances and the top up- and downregulated genes as well as principal component analysis plots. Figure 3. Examples of provided overview plots. 3A. MA-plot visualizing the distribution of differentially expressed genes by plotting the mean expression against the log fold change. 3B. Principle Component Analysis plot to visualize sample clustering. 3C. Heatmap showing sample to sample distances for given conditions. 3D. Heatmap displaying the 30 most upregulated genes. A similar heatmap is provided for the most downregulated genes, both allowing an easy identification of puta - tive candidate genes. A B C D Microsynth AG, SwitzerlandSch?tzenstrasse 15 ? P.O. Box ? CH - 9436 Balgach ? Phone + 41ff71ff722ff83 33 ? Fax + 41ff71ff722ff87 58 ? info @microsynth.ch ? www.microsynth.com THE SWISS DNA COMPANY Application Note ? Next Generation Sequencing References [1] Patro, R., Duggal, G., Love, M. I., Irizarry, R. A., & Kingsford, C. (2017). Salmon provides fast and bias-aware quantification of transcript expression. Nature Methods. [2] Dobin et al. (2013) STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21. https://doi.org/10.1093/bioinformatics/bts635 [3] Love et al. (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15. https://doi.org/10.1186/s13059-014- 0550-8 The results for the complementary pathway analysis include graphical rep - resentations of the up- and downreg - ulated pathways for both gene ontol - ogy and KEGG terms (see Figure 4 ). The pathway analysis module is available for most common model organisms such as human, mouse, rat or E. coli . Information is provided for all genes involved in a pathway count and log fold change, allowing a more detailed analysis of the identified pathways. Figure 4. Graphical output of the pathway analysis module. 4A. Excerpt from a network graph for upregulated pathways where colored nodes represent significantly upregulated gene ontology terms. 4B. Detail of a KEGG pathway analysis result. Both are available for up- and downregulated pathways. Related Services For the analysis of the small RNA frac - tion (e.g. microRNA) of an organism, please refer to our small RNA sequenc - ing service. If you are interested in the transcriptional analysis of whole micro - bial communities, our shotgun meta - transcriptomics service offers an attrac - tive solution. A B Microsynth AG, SwitzerlandSch?tzenstrasse 15 ? P.O. Box ? CH - 9436 Balgach ? Phone + 41fl71fl722fl83 33 ? Fax + 41fl71fl722fl87 58 ? info @microsynth.ch ? www.microsynth.com THE SWISS DNA COMPANY Application Note ? Next Generation Sequencing