Single-cell transcriptomics#Single-cell RNA-seq

The development of high-throughput RNA sequencing (RNA-seq) and microarrays has made gene expression analysis a routine. RNA analysis was previously limited to tracing individual transcripts by Northern blots or quantitative PCR. Higher throughput and speed allow researchers to frequently characterize the expression profiles of populations of thousands of cells. The data from bulk assays has led to identifying genes differentially expressed in distinct cell populations, and biomarker discovery.{{Cite book |last=Szabo |first=David T. |title=Biomarkers in Toxicology |publisher=Academic Press |year=2014 |isbn=9780124046306 |pages=1033–1038 |chapter=Chapter 62 - Transcriptomic biomarkers in safety and risk assessment of chemicals}}File:RNA Seq Experiment.pngThese studies are limited as they provide measurements for whole tissues and, as a result, show an average expression profile for all the constituent cells. This has a couple of drawbacks. Firstly, different cell types within the same tissue can have distinct roles in multicellular organisms. They often form subpopulations with unique transcriptional profiles. Correlations in the gene expression of the subpopulations can often be missed due to the lack of subpopulation identification. Secondly, bulk assays fail to recognize whether a change in the expression profile is due to a change in regulation or composition — for example if one cell type arises to dominate the population. Lastly, when your goal is to study cellular progression through differentiation, average expression profiles can only order cells by time rather than by developmental stage. Consequently, they cannot show trends in gene expression levels specific to certain stages.{{Cite journal |last=Trapnell |first=Cole |date=October 2015 |title=Defining cell types and states with single-cell genomics |journal=Genome Research |volume=25 |issue=10 |pages=1491–1498 |doi=10.1101/gr.190595.115 |issn=1549-5469 |pmc=4579334 |pmid=26430159}}

Recent advances in biotechnology allow the measurement of gene expression in hundreds to thousands of individual cells simultaneously. While these breakthroughs in transcriptomics technologies have enabled the generation of single-cell transcriptomic data, they also presented new computational and analytical challenges. Bioinformaticians can use techniques from bulk RNA-seq for single-cell data. Still, many new computational approaches have had to be designed for this data type to facilitate a complete and detailed study of single-cell expression profiles.{{Cite journal |last1=Stegle |first1=O. |last2=Teichmann |first2=S. |last3=Marioni |first3=J. |date=2015 |title=Computational and analytical challenges in single-cell transcriptomics |journal=Nature Reviews Genetics |volume=16 |issue=3 |pages=133–145 |doi=10.1038/nrg3833|pmid=25628217 |s2cid=205486032 }}

There is so far no standardized technique to generate single-cell data: all methods must include cell isolation from the population, lysate formation, amplification through reverse transcription, and quantification of expression levels. Common techniques for measuring expression are quantitative PCR or RNA-seq.{{cite journal |last1=Kolodziejczyk |first1=Aleksandra A. |last2=Kim |first2=Jong Kyoung |last3=Svensson |first3=Valentine |last4=Marioni |first4=John C. |last5=Teichmann |first5=Sarah A. |title=The Technology and Biology of Single-Cell RNA Sequencing |journal=Molecular Cell |date=May 2015 |volume=58 |issue=4 |pages=610–620 |doi=10.1016/j.molcel.2015.04.005|pmid=26000846 |doi-access=free}}

File:Fluorescence Assisted Cell Sorting (FACS) B2.jpg

Several methods are available to isolate and amplify cells for single-cell analysis, differing primarily in throughput and potential for cell selection. Low-throughput techniques, such as micropipetting, cytoplasmic aspiration,{{Cite web |title=Wayback Machine |url=http://www.single-cell-analysis.com/tag/cytoplasmic-aspiration/ |archive-url=http://web.archive.org/web/20231130011025/http://www.single-cell-analysis.com/tag/cytoplasmic-aspiration/ |archive-date=2023-11-30 |access-date=2025-04-15 |website=www.single-cell-analysis.com}} and laser capture microdissection, typically isolate hundreds of cells but enable deliberate cell selection.

High-throughput methods allow for the rapid isolation of hundreds to tens of thousands of cells.{{cite journal |last1=Poulin |first1=Jean-Francois |last2=Tasic |first2=Bosiljka |last3=Hjerling-Leffler |first3=Jens |last4=Trimarchi |first4=Jeffrey M. |last5=Awatramani |first5=Rajeshwar |date=1 September 2016 |title=Disentangling neural cell diversity using single-cell transcriptomics |journal=Nature Neuroscience |language=en |volume=19 |issue=9 |pages=1131–1141 |doi=10.1038/nn.4366 |issn=1097-6256 |pmid=27571192 |s2cid=14461377}} Common high-throughput approaches include Fluorescence Activated Cell Sorting (FACS) and the use of microfluidic devices. Microfluidic platforms often isolate single cells either by mechanical separation into microwells (e.g., BD Rhapsody, Takara ICELL8, Vycap Puncher Platform, CellMicrosystems CellRaft) or by encapsulation within droplets (e.g., 10x Genomics Chromium, Illumina Bio-Rad ddSEQ, 1CellBio InDrop, Dolomite Bio Nadia).{{cite journal |vauthors=Valihrach L, Androvic P, Kubista M |date=March 2018 |title=Platforms for Single-Cell Collection and Analysis |journal=International Journal of Molecular Sciences |volume=19 |issue=3 |page=807 |doi=10.3390/ijms19030807 |pmc=5877668 |pmid=29534489 |doi-access=free}} Furthermore, optimized protocols have been developed by integrating these isolation techniques directly with scRNA-seq workflows. For instance, combining FACS with scRNA-seq led to protocols like SORT-seq,{{Cite journal |last1=Muraro |first1=Mauro J. |last2=Dharmadhikari |first2=Gitanjali |last3=Grün |first3=Dominic |last4=Groen |first4=Nathalie |last5=Dielen |first5=Tim |last6=Jansen |first6=Erik |last7=van Gurp |first7=Leon |last8=Engelse |first8=Marten A. |last9=Carlotti |first9=Francoise |last10=de Koning |first10=Eelco J. P. |last11=van Oudenaarden |first11=Alexander |date=2016-10-26 |title=A Single-Cell Transcriptome Atlas of the Human Pancreas |journal=Cell Systems |volume=3 |issue=4 |pages=385–394.e3 |doi=10.1016/j.cels.2016.09.002 |issn=2405-4712 |pmc=5092539 |pmid=27693023}} and a list of studies utilizing SORT-seq can be found here.{{Cite web |title=SORT-seq Archives |url=https://www.scdiscoveries.com/publications/services/sort-seq-publications/ |access-date=2022-11-15 |website=Single Cell Discoveries |language=en-US}} Similarly, the integration of microfluidic devices with scRNA-seq has been highly optimized in protocols such as those developed by 10x Genomics.{{Cite journal |last1=Zheng |first1=Grace X. Y. |last2=Terry |first2=Jessica M. |last3=Belgrader |first3=Phillip |last4=Ryvkin |first4=Paul |last5=Bent |first5=Zachary W. |last6=Wilson |first6=Ryan |last7=Ziraldo |first7=Solongo B. |last8=Wheeler |first8=Tobias D. |last9=McDermott |first9=Geoff P. |last10=Zhu |first10=Junjie |last11=Gregory |first11=Mark T. |last12=Shuga |first12=Joe |last13=Montesclaros |first13=Luz |last14=Underwood |first14=Jason G. |last15=Masquelier |first15=Donald A. |date=2017-01-16 |title=Massively parallel digital transcriptional profiling of single cells |journal=Nature Communications |language=en |volume=8 |issue=1 |pages=14049 |doi=10.1038/ncomms14049 |pmid=28091601 |pmc=5241818 |bibcode=2017NatCo...814049Z |issn=2041-1723|doi-access=free }}

Single cells are labeled by adding beads with barcoded oligonucleotides; both cells and beads are supplied in limited amounts such that co-occupancy with multiple cells and beads is a very rare event.

To measure the level of expression of each transcript qPCR can be applied. Gene specific primers are used to amplify the corresponding gene as with regular PCR and as a result data is usually only obtained for sample sizes of less than 100 genes. The inclusion of housekeeping genes, whose expression should be constant under the conditions, is used for normalization. The most commonly used house keeping genes include GAPDH and α-actin, although the reliability of normalization through this process is questionable as there is evidence that the level of expression can vary significantly.{{cite journal|last1=Radonić|first1=Aleksandar|last2=Thulke|first2=Stefanie|last3=Mackay|first3=Ian M.|last4=Landt|first4=Olfert|last5=Siegert|first5=Wolfgang|last6=Nitsche|first6=Andreas|title=Guideline to reference gene selection for quantitative real-time PCR|journal=Biochemical and Biophysical Research Communications|date=23 January 2004|volume=313|issue=4|pages=856–862|pmid=14706621|issn=0006-291X|doi=10.1016/j.bbrc.2003.11.177}} Fluorescent dyes are used as reporter molecules to detect the PCR product and monitor the progress of the amplification - the increase in fluorescence intensity is proportional to the amplicon concentration. A plot of fluorescence vs. cycle number is made and a threshold fluorescence level is used to find cycle number at which the plot reaches this value. The cycle number at this point is known as the threshold cycle (C_t) and is measured for each gene.{{cite journal|last1=Wildsmith|first1=S. E.|last2=Archer|first2=G. E.|last3=Winkley|first3=A. J.|last4=Lane|first4=P. W.|last5=Bugelski|first5=P. J.|title=Maximization of signal derived from cDNA microarrays|journal=BioTechniques|date=1 January 2001|volume=30|issue=1|pages=202–206, 208|pmid=11196312|issn=0736-6205|doi=10.2144/01301dd04|doi-access=free}}

The single-cell RNA-seq technique converts a population of RNAs to a library of cDNA fragments. Single cells are labeled by adding beads with barcoded oligonucleotides; both cells and beads are supplied in limited amounts such that co-occupancy with multiple cells and beads is a very rare event. Once reverse transcription is complete, the cDNAs from many cells can be mixed together for sequencing.

File:RNA-Seq_workflow-5.pdf

These fragments are sequenced by high-throughput next generation sequencing techniques and the reads are mapped back to the reference genome, providing a count of the number of reads associated with each gene.{{cite journal |last1=Wang |first1=Zhong |last2=Gerstein |first2=Mark |last3=Snyder |first3=Michael |title=RNA-Seq: a revolutionary tool for transcriptomics |journal=Nature Reviews. Genetics |date=23 March 2017 |volume=10 |issue=1 |pages=57–63 |doi=10.1038/nrg2484 |pmid=19015660 |pmc=2949280 |issn=1471-0056}} Transcripts from a particular cell are identified by each cell's unique barcode.{{cite journal |vauthors=Klein AM, Mazutis L, Akartuna I, Tallapragada N, Veres A, Li V, Peshkin L, Weitz DA, Kirschner MW |date=May 2015 |title=Droplet barcoding for single-cell transcriptomics applied to embryonic stem cells |journal=Cell |volume=161 |issue=5 |pages=1187–1201 |doi=10.1016/j.cell.2015.04.044 |pmc=4441768 |pmid=26000487}}{{cite journal |vauthors=Macosko EZ, Basu A, Satija R, Nemesh J, Shekhar K, Goldman M, Tirosh I, Bialas AR, Kamitaki N, Martersteck EM, Trombetta JJ, Weitz DA, Sanes JR, Shalek AK, Regev A, McCarroll SA |date=May 2015 |title=Highly Parallel Genome-wide Expression Profiling of Individual Cells Using Nanoliter Droplets |journal=Cell |volume=161 |issue=5 |pages=1202–1214 |doi=10.1016/j.cell.2015.05.002 |pmc=4481139 |pmid=26000488}}

Normalization of RNA-Seq data accounts for cell to cell variation in the efficiencies of the cDNA library formation and sequencing. One method relies on the use of extrinsic RNA spike-ins that are added in equal quantities to each cell lysate and used to normalize read count by the number of reads mapped to spike-in mRNA.{{cite journal|last1=Jiang|first1=Lichun|last2=Schlesinger|first2=Felix|last3=Davis|first3=Carrie A.|last4=Zhang|first4=Yu|last5=Li|first5=Renhua|last6=Salit|first6=Marc|last7=Gingeras|first7=Thomas R.|last8=Oliver|first8=Brian|title=Synthetic spike-in standards for RNA-seq experiments|journal=Genome Research|date=23 March 2017|volume=21|issue=9|pages=1543–1551|doi=10.1101/gr.121095.111|pmid=21816910|pmc=3166838|issn=1088-9051}} Another control uses unique molecular identifiers (UMIs)-short DNA sequences (6–10nt) that are added to each cDNA before amplification and act as a bar code for each cDNA molecule. Normalization is achieved by using the count number of unique UMIs associated with each gene to account for differences in amplification efficiency.{{cite journal|last1=Islam|first1=Saiful|last2=Zeisel|first2=Amit|last3=Joost|first3=Simon|last4=La Manno|first4=Gioele|last5=Zajac|first5=Pawel|last6=Kasper|first6=Maria|last7=Lönnerberg|first7=Peter|last8=Linnarsson|first8=Sten|title=Quantitative single-cell RNA-seq with unique molecular identifiers|journal=Nature Methods|date=1 February 2014|volume=11|issue=2|pages=163–166|doi=10.1038/nmeth.2772|pmid=24363023|s2cid=6765530|language=en|issn=1548-7091}}

A combination of both spike-ins, UMIs and other approaches have been combined to help identify artifacts during library preparation{{cite journal |vauthors=Islam S, Zeisel A, Joost S, La Manno G, Zajac P, Kasper M, Lönnerberg P, Linnarsson S |date=February 2014 |title=Quantitative single-cell RNA-seq with unique molecular identifiers |journal=Nature Methods |volume=11 |issue=2 |pages=163–6 |doi=10.1038/nmeth.2772 |pmid=24363023 |s2cid=6765530}} and for more accurate normalization.

scRNA-Seq is becoming widely used across biological disciplines including Development, Neurology,{{cite journal |vauthors=Raj B, Wagner DE, McKenna A, Pandey S, Klein AM, Shendure J, Gagnon JA, Schier AF |date=June 2018 |title=Simultaneous single-cell profiling of lineages and cell types in the vertebrate brain |journal=Nature Biotechnology |volume=36 |issue=5 |pages=442–450 |doi=10.1038/nbt.4103 |pmc=5938111 |pmid=29608178}} Oncology,{{cite journal |vauthors=Olmos D, Arkenau HT, Ang JE, Ledaki I, Attard G, Carden CP, Reid AH, A'Hern R, Fong PC, Oomen NB, Molife R, Dearnaley D, Parker C, Terstappen LW, de Bono JS |date=January 2009 |title=Circulating tumour cell (CTC) counts as intermediate end points in castration-resistant prostate cancer (CRPC): a single-centre experience |journal=Annals of Oncology |volume=20 |issue=1 |pages=27–33 |doi=10.1093/annonc/mdn544 |pmid=18695026 |doi-access=free}}{{cite journal |vauthors=Levitin HM, Yuan J, Sims PA |date=April 2018 |title=Single-Cell Transcriptomic Analysis of Tumor Heterogeneity |url= |journal=Trends in Cancer |language=en |volume=4 |issue=4 |pages=264–268 |doi=10.1016/j.trecan.2018.02.003 |pmc=5993208 |pmid=29606308}}{{cite journal |vauthors=Jerby-Arnon L, Shah P, Cuoco MS, Rodman C, Su MJ, Melms JC, Leeson R, Kanodia A, Mei S, Lin JR, Wang S, Rabasha B, Liu D, Zhang G, Margolais C, Ashenberg O, Ott PA, Buchbinder EI, Haq R, Hodi FS, Boland GM, Sullivan RJ, Frederick DT, Miao B, Moll T, Flaherty KT, Herlyn M, Jenkins RW, Thummalapalli R, Kowalczyk MS, Cañadas I, Schilling B, Cartwright AN, Luoma AM, Malu S, Hwu P, Bernatchez C, Forget MA, Barbie DA, Shalek AK, Tirosh I, Sorger PK, Wucherpfennig K, Van Allen EM, Schadendorf D, Johnson BE, Rotem A, Rozenblatt-Rosen O, Garraway LA, Yoon CH, Izar B, Regev A |date=November 2018 |title=A Cancer Cell Program Promotes T Cell Exclusion and Resistance to Checkpoint Blockade |url= |journal=Cell |language=en |volume=175 |issue=4 |pages=984–997.e24 |doi=10.1016/j.cell.2018.09.006 |pmc=6410377 |pmid=30388455}} Autoimmune disease,{{cite journal |vauthors=Stephenson W, Donlin LT, Butler A, Rozo C, Bracken B, Rashidfarrokhi A, Goodman SM, Ivashkiv LB, Bykerk VP, Orange DE, Darnell RB, Swerdlow HP, Satija R |date=February 2018 |title=Single-cell RNA-seq of rheumatoid arthritis synovial tissue using low-cost microfluidic instrumentation |journal=Nature Communications |volume=9 |issue=1 |pages=791 |bibcode=2018NatCo...9..791S |doi=10.1038/s41467-017-02659-x |pmc=5824814 |pmid=29476078}} and Infectious disease.{{cite journal |vauthors=Avraham R, Haseley N, Brown D, Penaranda C, Jijon HB, Trombetta JJ, Satija R, Shalek AK, Xavier RJ, Regev A, Hung DT |date=September 2015 |title=Pathogen Cell-to-Cell Variability Drives Heterogeneity in Host Immune Responses |journal=Cell |volume=162 |issue=6 |pages=1309–21 |doi=10.1016/j.cell.2015.08.027 |pmc=4578813 |pmid=26343579}} Several scRNA-Seq protocols have been published: Tang et al.,{{cite journal |vauthors=Tang F, Barbacioru C, Wang Y, Nordman E, Lee C, Xu N, Wang X, Bodeau J, Tuch BB, Siddiqui A, Lao K, Surani MA |date=May 2009 |title=mRNA-Seq whole-transcriptome analysis of a single cell |journal=Nature Methods |volume=6 |issue=5 |pages=377–82 |doi=10.1038/NMETH.1315 |pmid=19349980 |s2cid=16570747}} STRT,{{cite journal |vauthors=Islam S, Kjällquist U, Moliner A, Zajac P, Fan JB, Lönnerberg P, Linnarsson S |date=July 2011 |title=Characterization of the single-cell transcriptional landscape by highly multiplex RNA-seq |journal=Genome Research |volume=21 |issue=7 |pages=1160–7 |doi=10.1101/gr.110882.110 |pmc=3129258 |pmid=21543516}} SMART-seq,{{cite journal |vauthors=Ramsköld D, Luo S, Wang YC, Li R, Deng Q, Faridani OR, Daniels GA, Khrebtukova I, Loring JF, Laurent LC, Schroth GP, Sandberg R |date=August 2012 |title=Full-length mRNA-Seq from single-cell levels of RNA and individual circulating tumor cells |journal=Nature Biotechnology |volume=30 |issue=8 |pages=777–82 |doi=10.1038/nbt.2282 |pmc=3467340 |pmid=22820318}} CEL-seq,{{cite journal |vauthors=Hashimshony T, Wagner F, Sher N, Yanai I |date=September 2012 |title=CEL-Seq: single-cell RNA-Seq by multiplexed linear amplification |journal=Cell Reports |volume=2 |issue=3 |pages=666–73 |doi=10.1016/j.celrep.2012.08.003 |pmid=22939981 |doi-access=free}} RAGE-seq,{{cite journal |vauthors=Singh M, Al-Eryani G, Carswell S, Ferguson JM, Blackburn J, Barton K, Roden D, Luciani F, Phan T, Junankar S, Jackson K, Goodnow CC, Smith MA, Swarbrick A |year=2018 |title=High-throughput targeted long-read single cell sequencing reveals the clonal and transcriptional landscape of lymphocytes |journal=bioRxiv |volume=10 |issue=1 |page=3120 |doi=10.1101/424945 |pmc=6635368 |pmid=31311926 |doi-access=free}} Quartz-seq{{cite journal |vauthors=Sasagawa Y, Nikaido I, Hayashi T, Danno H, Uno KD, Imai T, Ueda HR |date=April 2013 |title=Quartz-Seq: a highly reproducible and sensitive single-cell RNA sequencing method, reveals non-genetic gene-expression heterogeneity |journal=Genome Biology |volume=14 |issue=4 |pages=R31 |doi=10.1186/gb-2013-14-4-r31 |pmc=4054835 |pmid=23594475 |doi-access=free}} and C1-CAGE.{{cite journal |vauthors=Kouno T, Moody J, Kwon AT, Shibayama Y, Kato S, Huang Y, Böttcher M, Motakis E, Mendez M, Severin J, Luginbühl J, Abugessaisa I, Hasegawa A, Takizawa S, Arakawa T, Furuno M, Ramalingam N, West J, Suzuki H, Kasukawa T, Lassmann T, Hon CC, Arner E, Carninci P, Plessy C, Shin JW |date=January 2019 |title=C1 CAGE detects transcription start sites and enhancer activity at single-cell resolution |journal=Nature Communications |volume=10 |issue=1 |pages=360 |bibcode=2019NatCo..10..360K |doi=10.1038/s41467-018-08126-5 |pmc=6341120 |pmid=30664627}} These protocols differ in terms of strategies for reverse transcription, cDNA synthesis and amplification, and the possibility to accommodate sequence-specific barcodes (i.e. UMIs) or the ability to process pooled samples.{{cite journal |vauthors=Dal Molin A, Di Camillo B |year=2019 |title=How to design a single-cell RNA-sequencing experiment: pitfalls, challenges and perspectives |journal=Briefings in Bioinformatics |volume=20 |issue=4 |pages=1384–1394 |doi=10.1093/bib/bby007 |pmid=29394315}} In 2017, two approaches were introduced to simultaneously measure single-cell mRNA and protein expression through oligonucleotide-labeled antibodies known as REAP-seq,{{cite journal |first6=Lixia |first7=Jerelyn |first8=Namit |first9=Kelvin Xi |vauthors=Peterson VM, Zhang KX, Kumar N, Wong J, Li L, Wilson DC, Moore R, McClanahan TK, Sadekova S, Klappenbach JA |date=October 2017 |title=Multiplexed quantification of proteins and transcripts in single cells |journal=Nature Biotechnology |volume=35 |issue=10 |pages=936–939 |doi=10.1038/nbt.3973 |pmid=28854175 |s2cid=205285357}} and CITE-seq.{{cite journal |first5=Brian |first6=William |first7=Christoph |first8=Marlon |vauthors=Stoeckius M, Hafemeister C, Stephenson W, Houck-Loomis B, Chattopadhyay PK, Swerdlow H, Satija R, Smibert P |date=September 2017 |title=Simultaneous epitope and transcriptome measurement in single cells |journal=Nature Methods |volume=14 |issue=9 |pages=865–868 |doi=10.1038/nmeth.4380 |pmc=5669064 |pmid=28759029}}

scRNA-Seq has provided considerable insight into the development of embryos and organisms, including the worm Caenorhabditis elegans,{{cite journal |vauthors=Cao J, Packer JS, Ramani V, Cusanovich DA, Huynh C, Daza R, Qiu X, Lee C, Furlan SN, Steemers FJ, Adey A, Waterston RH, Trapnell C, Shendure J |date=August 2017 |title=Comprehensive single-cell transcriptional profiling of a multicellular organism |journal=Science |volume=357 |issue=6352 |pages=661–667 |bibcode=2017Sci...357..661C |doi=10.1126/science.aam8940 |pmc=5894354 |pmid=28818938}} and the regenerative planarian Schmidtea mediterranea.{{cite journal |vauthors=Plass M, Solana J, Wolf FA, Ayoub S, Misios A, Glažar P, Obermayer B, Theis FJ, Kocks C, Rajewsky N |date=May 2018 |title=Cell type atlas and lineage tree of a whole complex animal by single-cell transcriptomics |journal=Science |volume=360 |issue=6391 |pages=eaaq1723 |doi=10.1126/science.aaq1723 |pmid=29674432 |doi-access=free}}{{cite journal |vauthors=Fincher CT, Wurtzel O, de Hoog T, Kravarik KM, Reddien PW |date=May 2018 |title=Schmidtea mediterranea |journal=Science |volume=360 |issue=6391 |pages=eaaq1736 |doi=10.1126/science.aaq1736 |pmc=6563842 |pmid=29674431}} The first vertebrate animals to be mapped in this way were Zebrafish{{cite journal |vauthors=Wagner DE, Weinreb C, Collins ZM, Briggs JA, Megason SG, Klein AM |date=June 2018 |title=Single-cell mapping of gene expression landscapes and lineage in the zebrafish embryo |journal=Science |volume=360 |issue=6392 |pages=981–987 |bibcode=2018Sci...360..981W |doi=10.1126/science.aar4362 |pmc=6083445 |pmid=29700229}}{{cite journal |vauthors=Farrell JA, Wang Y, Riesenfeld SJ, Shekhar K, Regev A, Schier AF |date=June 2018 |title=Single-cell reconstruction of developmental trajectories during zebrafish embryogenesis |journal=Science |volume=360 |issue=6392 |pages=eaar3131 |doi=10.1126/science.aar3131 |pmc=6247916 |pmid=29700225}} and Xenopus laevis.{{cite journal |vauthors=Briggs JA, Weinreb C, Wagner DE, Megason S, Peshkin L, Kirschner MW, Klein AM |date=June 2018 |title=The dynamics of gene expression in vertebrate embryogenesis at single-cell resolution |journal=Science |volume=360 |issue=6392 |pages=eaar5780 |doi=10.1126/science.aar5780 |pmc=6038144 |pmid=29700227}} In each case multiple stages of the embryo were studied, allowing the entire process of development to be mapped on a cell-by-cell basis.{{cite journal |vauthors=Griffith M, Walker JR, Spies NC, Ainscough BJ, Griffith OL |date=August 2015 |title=Informatics for RNA Sequencing: A Web Resource for Analysis on the Cloud |journal=PLOS Computational Biology |volume=11 |issue=8 |pages=e1004393 |bibcode=2015PLSCB..11E4393G |doi=10.1371/journal.pcbi.1004393 |pmc=4527835 |pmid=26248053 |doi-access=free}} Science recognized these advances as the 2018 Breakthrough of the Year.{{cite web |title=Science's 2018 Breakthrough of the Year: tracking development cell by cell |url=https://vis.sciencemag.org/breakthrough2018/finalists/ |work=Science Magazine |publisher=American Association for the Advancement of Science |vauthors=You J}}

A problem associated with single-cell data occurs in the form of zero inflated gene expression distributions, known as technical dropouts, that are common due to low mRNA concentrations of less-expressed genes that are not captured in the reverse transcription process. The percentage of mRNA molecules in the cell lysate that are detected is often only 10-20%.{{cite journal|last1=Kharchenko|first1=Peter V.|last2=Silberstein|first2=Lev|last3=Scadden|first3=David T.|title=Bayesian approach to single-cell differential expression analysis|journal=Nature Methods|date=1 July 2014|volume=11|issue=7|pages=740–742|doi=10.1038/nmeth.2967|pmid=24836921|language=en|issn=1548-7091|pmc=4112276}}

When using RNA spike-ins for normalization the assumption is made that the amplification and sequencing efficiencies for the endogenous and spike-in RNA are the same. Evidence suggests that this is not the case given fundamental differences in size and features, such as the lack of a polyadenylated tail in spike-ins and therefore shorter length.{{cite journal|last1=Svensson|first1=Valentine|last2=Natarajan|first2=Kedar Nath|last3=Ly|first3=Lam-Ha|last4=Miragaia|first4=Ricardo J.|last5=Labalette|first5=Charlotte|last6=Macaulay|first6=Iain C.|last7=Cvejic|first7=Ana|last8=Teichmann|first8=Sarah A.|title=Power analysis of single-cell RNA-sequencing experiments|journal=Nature Methods|date=6 March 2017|volume=advance online publication|issue=4|pages=381–387|doi=10.1038/nmeth.4220|pmid=28263961|pmc=5376499|language=en|issn=1548-7105}} Additionally, normalization using UMIs assumes the cDNA library is sequenced to saturation, which is not always the case.

In the amplification step, either PCR or in vitro transcription (IVT) is currently used to amplify cDNA. One of the advantages of PCR-based methods is the ability to generate full-length cDNA. However, different PCR efficiency on particular sequences (for instance, GC content and snapback structure) may also be exponentially amplified, producing libraries with uneven coverage. On the other hand, while libraries generated by IVT can avoid PCR-induced sequence bias, specific sequences may be transcribed inefficiently, thus causing sequence drop-out or generating incomplete sequences.{{cite journal |vauthors=Eberwine J, Sul JY, Bartfai T, Kim J |date=January 2014 |title=The promise of single-cell sequencing |journal=Nature Methods |volume=11 |issue=1 |pages=25–7 |doi=10.1038/nmeth.2769 |pmid=24524134 |s2cid=11575439}}"{{cite journal |vauthors=Shapiro E, Biezuner T, Linnarsson S |date=September 2013 |title=Single-cell sequencing-based technologies will revolutionize whole-organism science |journal=Nature Reviews. Genetics |volume=14 |issue=9 |pages=618–30 |doi=10.1038/nrg3542 |pmid=23897237 |s2cid=500845}}"

Challenges for scRNA-Seq include preserving the initial relative abundance of mRNA in a cell and identifying rare transcripts."{{cite journal |vauthors=Hebenstreit D |date=November 2012 |title=Methods, Challenges and Potentials of Single Cell RNA-seq |journal=Biology |volume=1 |issue=3 |pages=658–67 |doi=10.3390/biology1030658 |pmc=4009822 |pmid=24832513 |doi-access=free}}" The reverse transcription step is critical as the efficiency of the RT reaction determines how much of the cell's RNA population will be eventually analyzed by the sequencer. The processivity of reverse transcriptases and the priming strategies used may affect full-length cDNA production and the generation of libraries biased toward the 3’ or 5' end of genes.

A further consideration when sequencing large, branched cell types, such as neurons, comes from the removal of distal processes containing local pools of RNA during the single-cell isolation process. In these cells, scRNA-seq datasets only capture transcript in the central cell body, omitting transcripts from RNA pools localized to cellular processes that can be involved in local translation or other RNA-mediated subcellular mechanisms. In the brain it has been estimated that over 40% of total RNA is not sequenced by scRNA-seq due to the prevalence of local transcriptomes in cellular processes such as axons, dendrites, myelin, and endfeet. {{cite journal | last=Ament | first=Seth A. | last2=Poulopoulos | first2=Alexandros | title=The brain's dark transcriptome: Sequencing RNA in distal compartments of neurons and glia | journal=Current Opinion in Neurobiology | volume=81 | date=2023 | pmid=37196598 | pmc=10524153 | doi=10.1016/j.conb.2023.102725 | doi-access=free | page=102725}}

Insights based on single-cell data analysis assume that the input is a matrix of normalized gene expression counts, generated by the approaches outlined above, and can provide opportunities that are not obtainable by bulk.

Three main insights provided:{{cite journal |last1=Stegle |first1=Oliver |last2=Teichmann |first2=Sarah A. |last3=Marioni |first3=John C. |date=1 March 2015 |title=Computational and analytical challenges in single-cell transcriptomics |journal=Nature Reviews Genetics |language=en |volume=16 |issue=3 |pages=133–145 |doi=10.1038/nrg3833 |issn=1471-0056 |pmid=25628217 |s2cid=205486032}}

1. Identification and characterization of cell types and their spatial organisation in time
2. Inference of gene regulatory networks and their strength across individual cells
3. Classification of the stochastic component of transcription

The techniques outlined have been designed to help visualise and explore patterns in the data in order to facilitate the revelation of these three features.

File:KMeans-Gaussian-data.svg

File:Iris dendrogram.png

Clustering allows for the formation of subgroups in the cell population. Cells can be clustered by their transcriptomic profile in order to analyse the sub-population structure and identify rare cell types or cell subtypes. Alternatively, genes can be clustered by their expression states in order to identify covarying genes. A combination of both clustering approaches, known as biclustering, has been used to simultaneously cluster by genes and cells to find genes that behave similarly within cell clusters.{{cite journal|last1=Buettner|first1=Florian|last2=Natarajan|first2=Kedar N.|last3=Casale|first3=F. Paolo|last4=Proserpio|first4=Valentina|last5=Scialdone|first5=Antonio|last6=Theis|first6=Fabian J.|last7=Teichmann|first7=Sarah A.|last8=Marioni|first8=John C.|last9=Stegle|first9=Oliver|title=Computational analysis of cell-to-cell heterogeneity in single-cell RNA-sequencing data reveals hidden subpopulations of cells|journal=Nature Biotechnology|date=1 February 2015|volume=33|issue=2|pages=155–160|doi=10.1038/nbt.3102|pmid=25599176|language=en|issn=1087-0156|doi-access=free}}

Clustering methods applied can be K-means clustering, forming disjoint groups or Hierarchical clustering, forming nested partitions.

Biclustering provides several advantages by improving the resolution of clustering. Genes that are only informative to a subset of cells and are hence only expressed there can be identified through biclustering. Moreover, similarly behaving genes that differentiate one cell cluster from another can be identified using this method.{{cite journal|last1=Ntranos|first1=Vasilis|last2=Kamath|first2=Govinda M.|last3=Zhang|first3=Jesse M.|last4=Pachter|first4=Lior|last5=Tse|first5=David N.|title=Fast and accurate single-cell RNA-seq analysis by clustering of transcript-compatibility counts|journal=Genome Biology|date=26 May 2016|volume=17|issue=1|pages=112|doi=10.1186/s13059-016-0970-8|pmid=27230763|pmc=4881296|issn=1474-7596 |doi-access=free }}

File:PCA of Guinean and other African populations Y chromosome haplogroup frequencies.jpg

Dimensionality reduction algorithms such as Principal component analysis (PCA) and t-SNE can be used to simplify data for visualisation and pattern detection by transforming cells from a high to a lower dimensional space. The result of this method produces graphs with each cell as a point in a 2-D or 3-D space. Dimensionality reduction is frequently used before clustering as cells in high dimensions can wrongly appear to be close due to distance metrics behaving non-intuitively.{{cite journal|last1=Pierson|first1=Emma|last2=Yau|first2=Christopher|title=ZIFA: Dimensionality reduction for zero-inflated single-cell gene expression analysis|journal=Genome Biology|date=1 January 2015|volume=16|pages=241|doi=10.1186/s13059-015-0805-z|pmid=26527291|pmc=4630968|issn=1474-760X |doi-access=free }}

The most frequently used technique is PCA, which identifies the directions of largest variance principal components and transforms the data so that the first principal component has the largest possible variance, and successive principle components in turn each have the highest variance possible while remaining orthogonal to the preceding components. The contribution each gene makes to each component is used to infer which genes are contributing the most to variance in the population and are involved in differentiating different subpopulations.{{cite journal|last1=Treutlein|first1=Barbara|last2=Brownfield|first2=Doug G.|last3=Wu|first3=Angela R.|last4=Neff|first4=Norma F.|last5=Mantalas|first5=Gary L.|last6=Espinoza|first6=F. Hernan|last7=Desai|first7=Tushar J.|last8=Krasnow|first8=Mark A.|last9=Quake|first9=Stephen R.|title=Reconstructing lineage hierarchies of the distal lung epithelium using single-cell RNA-seq|journal=Nature|date=15 May 2014|volume=509|issue=7500|pages=371–375|doi=10.1038/nature13173|pmid=24739965|language=en|pmc=4145853|bibcode=2014Natur.509..371T}}

Detecting differences in gene expression level between two populations is used both single-cell and bulk transcriptomic data. Specialised methods have been designed for single-cell data that considers single cell features such as technical dropouts and shape of the distribution e.g. Bimodal vs. unimodal.{{cite journal|last1=Korthauer|first1=Keegan D.|last2=Chu|first2=Li-Fang|last3=Newton|first3=Michael A.|last4=Li|first4=Yuan|last5=Thomson|first5=James|last6=Stewart|first6=Ron|last7=Kendziorski|first7=Christina|author7-link= Christina Kendziorski |title=A statistical approach for identifying differential distributions in single-cell RNA-seq experiments|journal=Genome Biology|date=1 January 2016|volume=17|issue=1|pages=222|doi=10.1186/s13059-016-1077-y|pmid=27782827|pmc=5080738|issn=1474-760X |doi-access=free }}

Gene ontology terms describe gene functions and the relationships between those functions into three classes:

1. Molecular function
2. Cellular component
3. Biological process

Gene Ontology (GO) term enrichment is a technique used to identify which GO terms are over-represented or under-represented in a given set of genes. In single-cell analysis input list of genes of interest can be selected based on differentially expressed genes or groups of genes generated from biclustering. The number of genes annotated to a GO term in the input list is normalized against the number of genes annotated to a GO term in the background set of all genes in genome to determine statistical significance.{{cite journal|last1=Haghverdi|first1=Laleh|last2=Büttner|first2=Maren|last3=Wolf|first3=F. Alexander|last4=Buettner|first4=Florian|last5=Theis|first5=Fabian J.|title=Diffusion pseudotime robustly reconstructs lineage branching|journal=Nature Methods|date=1 October 2016|volume=13|issue=10|pages=845–848|doi=10.1038/nmeth.3971|pmid=27571553|s2cid=3594049|language=en|issn=1548-7091|url=http://edoc.mdc-berlin.de/19027/1/19027oa.pdf}}

{{Main|Trajectory inference}}

File:7n graph with minimal spanning tree.svg

Pseudo-temporal ordering (or trajectory inference) is a technique that aims to infer gene expression dynamics from snapshot single-cell data. The method tries to order the cells in such a way that similar cells are closely positioned to each other. This trajectory of cells can be linear, but can also bifurcate or follow more complex graph structures. The trajectory, therefore, enables the inference of gene expression dynamics and the ordering of cells by their progression through differentiation or response to external stimuli.

The method relies on the assumptions that the cells follow the same path through the process of interest and that their transcriptional state correlates to their progression. The algorithm can be applied to both mixed populations and temporal samples.

More than 50 methods for pseudo-temporal ordering have been developed, and each has its own requirements for prior information (such as starting cells or time course data), detectable topologies, and methodology.{{cite journal|doi = 10.1101/276907| pages = 276907|last1 = Saelens| first1 = Wouter| last2 = Cannoodt| first2 = Robrecht| last3 = Todorov| first3 = Helena| last4 = Saeys| first4 = Yvan| title = A comparison of single-cell trajectory inference methods: towards more accurate and robust tools| journal = bioRxiv| accessdate = 2018-03-12| date = 2018-03-05|url = https://www.biorxiv.org/content/early/2018/03/05/276907| doi-access = free}} An example algorithm is the Monocle algorithm{{cite journal|last1=Trapnell|first1=Cole|last2=Cacchiarelli|first2=Davide|last3=Grimsby|first3=Jonna|last4=Pokharel|first4=Prapti|last5=Li|first5=Shuqiang|last6=Morse|first6=Michael|last7=Lennon|first7=Niall J.|last8=Livak|first8=Kenneth J.|last9=Mikkelsen|first9=Tarjei S.|last10=Rinn|first10=John L.|title=Pseudo-temporal ordering of individual cells reveals dynamics and regulators of cell fate decisions|journal=Nature Biotechnology|date=23 March 2017|volume=32|issue=4|pages=381–386|doi=10.1038/nbt.2859|pmid=24658644|pmc=4122333|issn=1087-0156}} that carries out dimensionality reduction of the data, builds a minimal spanning tree using the transformed data, orders cells in pseudo-time by following the longest connected path of the tree and consequently labels cells by type. Another example is the diffusion pseudotime (DPT) algorithm, which uses a diffusion map and diffusion process. Another class of methods such as MARGARET {{Cite journal|last1=Pandey|first1=Kushagra|last2=Zafar|first2=Hamim|date=2022|title=Inference of cell state transitions and cell fate plasticity from single-cell with MARGARET|journal=Nucleic Acids Research|volume=50 |issue=15 |pages=e86 |language=en|doi=10.1093/nar/gkac412|pmid=35639499 |pmc=9410915 |issn=0305-1048|doi-access=free}} employ graph partitioning for capturing complex trajectory topologies such as disconnected and multifurcating trajectories.

Gene regulatory network inference is a technique that aims to construct a network, shown as a graph, in which the nodes represent the genes and edges indicate co-regulatory interactions. The method relies on the assumption that a strong statistical relationship between the expression of genes is an indication of a potential functional relationship.{{cite book|last1=Wei|first1=J.|last2=Hu|first2=X.|last3=Zou|first3=X.|last4=Tian|first4=T.|title=2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) |chapter=Inference of genetic regulatory network for stem cell using single cells expression data |date=1 December 2016|pages=217–222|doi=10.1109/BIBM.2016.7822521|isbn=978-1-5090-1611-2|s2cid=27737735}} The most commonly used method to measure the strength of a statistical relationship is correlation. However, correlation fails to identify non-linear relationships and mutual information is used as an alternative. Gene clusters linked in a network signify genes that undergo coordinated changes in expression.{{cite journal|last1=Moignard|first1=Victoria|last2=Macaulay|first2=Iain C.|last3=Swiers|first3=Gemma|last4=Buettner|first4=Florian|last5=Schütte|first5=Judith|last6=Calero-Nieto|first6=Fernando J.|last7=Kinston|first7=Sarah|last8=Joshi|first8=Anagha|last9=Hannah|first9=Rebecca|last10=Theis|first10=Fabian J.|last11=Jacobsen|first11=Sten Eirik|last12=de Bruijn|first12=Marella F.|last13=Göttgens|first13=Berthold|title=Characterization of transcriptional networks in blood stem and progenitor cells using high-throughput single-cell gene expression analysis|journal=Nature Cell Biology|date=1 April 2013|volume=15|issue=4|pages=363–372|doi=10.1038/ncb2709|pmid=23524953|language=en|issn=1465-7392|pmc=3796878}}

The presence or strength of technical effects and the types of cells observed often differ in single-cell transcriptomics datasets generated using different experimental protocols and under different conditions. This difference results in strong batch effects that may bias the findings of statistical methods applied across batches, particularly in the presence of confounding.{{cite journal|last1=Hicks|first1=Stephanie C|last2=Townes|first2=William F|last3=Teng|first3=Mingxiang|last4=Irizarry|first4=Rafael A|title=Missing data and technical variability in single-cell RNA-sequencing experiments|journal=Biostatistics|date=6 November 2017|volume=19|issue=4|pages=562–578|doi=10.1093/biostatistics/kxx053|pmid=29121214|pmc=6215955|language=en}}

As a result of the aforementioned properties of single-cell transcriptomic data, batch correction methods developed for bulk sequencing data were observed to perform poorly. Consequently, researchers developed statistical methods to correct for batch effects that are robust to the properties of single-cell transcriptomic data to integrate data from different sources or experimental batches. Laleh Haghverdi performed foundational work in formulating the use of mutual nearest neighbors between each batch to define batch correction vectors.{{cite journal|last1=Haghverdi|first1=Laleh|last2=Lun|first2=Aaron T L|last3=Morgan|first3=Michael D|last4=Marioni|first4=John C|title=Batch effects in single-cell RNA-sequencing data are corrected by matching mutual nearest neighbors|journal=Nature Biotechnology|date=2 April 2018|volume=36|issue=5|pages=421–427|doi=10.1038/nbt.4091|pmid=29608177|pmc=6152897|language=en}} With these vectors, you can merge datasets that each include at least one shared cell type. An orthogonal approach involves the projection of each dataset onto a shared low-dimensional space using canonical correlation analysis.{{cite journal|last1=Butler|first1=Andrew|last2=Hoffman|first2=Paul|last3=Smibert|first3=Peter|last4=Papalexi|first4=Efthymia|last5=Satija|first5=Rahul|title=Integrating single-cell transcriptomic data across different conditions, technologies, and species|journal=Nature Biotechnology|date=2 April 2018|volume=36|issue=5|pages=421–427|doi=10.1038/nbt.4096|pmid=29608179|pmc=6700744|language=en}} Mutual nearest neighbors and canonical correlation analysis have also been combined to define integration "anchors" comprising reference cells in one dataset, to which query cells in another dataset are normalized.{{cite journal|last1=Stuart|first1=Tim|last2=Butler|first2=Andrew|last3=Hoffman|first3=Paul|last4=Hafemeister|first4=Christoph|last5=Papalexia|first5=Efthymia|last6=Mauck|first6=William M III|last7=Hao|first7=Yuhan|last8=Marlon|first8=Stoeckius|last9=Smibert|first9=Peter|last10=Satija|first10=Rahul|title=Comprehensive Integration of Single-Cell Data|journal=Cell|date=6 June 2019|volume=177|issue=7|pages=1888–1902|doi=10.1016/j.cell.2019.05.031|pmid=31178118|pmc=6687398|language=en}} Another class of methods (e.g., scDREAMER{{cite journal|last1=Shree|first1=Ajita|last2=Pavan|first2=Musale Krushna|last3=Zafar|first3=Hamim|title=scDREAMER for atlas-level integration of single-cell datasets using deep generative model paired with adversarial classifier|journal=Nature Communications|date=27 November 2023|volume=14|issue=1|pages=7781|doi=10.1038/s41467-023-43590-8|pmid= 38012145|language=en|doi-access=free|pmc=10682386 |bibcode=2023NatCo..14.7781S }}) uses deep generative models such as variational autoencoders for learning batch-invariant latent cellular representations which can be used for downstream tasks such as cell type clustering, denoising of single-cell gene expression vectors and trajectory inference.

• RNA-Seq
• Single-cell analysis
• Single-cell sequencing
• Transcriptome
• Transcriptomics

{{Reflist|35em}}

• [https://www.youtube.com/watch?v=2-N3SPPHt6o Dissecting Tumor Heterogeneity with Single-Cell Transcriptomics]
• [https://www.scdiscoveries.com/single-cell-sequencing-ultimate-guide/ The ultimate single-cell RNA sequencing guide] by single-cell RNA sequencing service provider Single Cell Discoveries.

Category:DNA sequencing

Category:Molecular biology techniques

Category:Biotechnology