Single Cell Expression Profiling

Just another WordPress site

Single Cell Expression Profiling

let me start by giving a brief background of Tata bio center we’ve been working in the pcr field since 1991 in the early days we developed eyes and probes that became four stand up combining DNA at that time we’d up did not know how to use them but that became obvious when Rossi Gucci invented PCR or qPCR in 1993 since then we have established collaborations with all leading QP Co instrument and solution providers running today the best equipped laboratories for qPCR work Europe we offer hands-on training in qPCR every month tarah organizes two to three courses somewhere around the world training annually in more than 700 researchers we are Europe’s leading provider of qPCR services offering routine analysis as well as latest technologies such as the RNase H dependent PCR for a mutation analysis recently developed by IDT we develop products for qPCR focusing on experimental design optimization and quality control and researchers and students are welcome to tata bringing their samples and performing experiments under our guidance using latest and most advanced qpcr tools and technologies taras member of the europeans pedia consortium that aims at standardizing the pre analytical process of molecular diagnostics among our activities we organize proficiency rain trials any European laboratory is welcome to join free of charge they receive samples from us with instructions strain trial we distributed blood samples and asked participating laboratories to extract RNA now return it to us for quality assessment we tested several quality parameters and found that’s 33% of the participating laboratories had at least two quality parameters out of control Tariq contributed to the writing of the Mikey guidelines advising qpcr workers how to perform and report qpc results such that quality can be assessed by independent researchers and results rep reduced the mikey guidelines are receiving great attention among qPCR users as well as journals tarah was also the first laboratory in europe to obtain flexible iso 17025 certification for nucleic acid analysis was qPCR but let’s talk about science this is an island of liang hands stained for alpha cells now the eyelid is stained also for alpha cells and here it is stained also for beta cells as seen the eyelid is a mosaic of the three cell types the three cell types are very different being involved in diverse functions and responding to different stimuli if the analyzed entire islet or any part of it we will measure the collective response of all the cells present rather than the response of any particular cell ty as a consequence the response of the interesting cell type may be swamped or blurred by the responses of the other non relevant cells to pro the complexity often encountered when studying biological samples we developed single-cell qpcr expression profiling this slide shows qpcr response curves of insulin 1 in individual meter cells collected from cell line the good news are the data are of excellent quality indeed the quality of single cell expression data is typically superior to the quality of conventional samples composed of many cells this may surprise reason is the technical issues of qPCR typically inhibition and RNA degradation are caused by extrasolar factors analyzing carefully washed single cells we rarely experience RNA degradation or PCR inhibition however there is substantial cell to cell variability even between seemingly like cells such as these collected from culture we find

large spread of qPCR response curtains summarizing the data and a histogram showing the frequency of cells containing number of transcript in certain ranges we find highly skewed distribution these data show expression of beat actin some 40 cells have between 0 and 100 transcripts 25 cells had between 100 and 200 transcripts and 15 cells had between 200 and 300 transcripts and so forth most cells Harbor very few transcripts only very very few cells are rich in transcripts these are the same data presented with logarithmic scale for the x-axis indicating the number of transcripts we see the spread can now be modeled was a normal distribution evidencing the distribution of beta actin transcripts among beta cell culture is consistent with a log normal distribution today we know log normal distributions of transcripts along like sounds is the norm being observed for most cell types as well as genes the log normal distribution has interesting consequences when we study classical samples based on thousands of cells we could measure total expression and divide was a number of cells to obtain an average this average known as the arithmetic average will not give the count of transcripts in the typical cell the typical cell is considered to be the cell in the middle if we take all the cell’s in the population and sort them based on the numbers of the particular transcript they contain because of the underlying log normal distribution the number of transcripts in the typical cell is given by the geometric average rather than the arithmetic average the geometric average of transcripts per cell cannot be determined in the classical experiment it can only be calculated from single cell measurements did you find that confusing well you are not alone this is a summary of our paper in Nature Genetics so what is the origin of the log normal distribution this has been shown in beautiful work by Jonathan Chubb and others using fluorescence in situ hybridization they measured expression in lives single cells over time let me show a movie the bright spot that appears disappears and reappears again our transcripts clearly the number of transcripts in the cell vary over time this is known as transcriptional bursting transcripts of a gene are produced on mass in bursts followed by slow decay then there is a new burst at a DK integrating over time we obtain a gamma distribution that is similar in shape to the log normal distribution we find for the distribution of transcripts among individual cells the pulse interval and pulse duration depends on gene activity and locus but are typically of the order of minutes to hours in our first study we measured expression of five genes per cell that was neat but not good enough we wanted to measure more genes and developed a highly optimized workflow for high-throughput single cell profiling this included fluorescence activated cell sorting of single cells direct lysis using a reagent compatible with downstream RT q PCR to eliminate losses due to washing highly optimized reverse transcription and pre amplification pre amplification is a highly multiplex PCR performed a limited number of cycles the purpose is to increase the concentration a number of target copies for downstream

single sec singer plaques qPCR each singer plaques qPCR should start with at least some 25 target copies to avoid sampling ambiguity a limited number of amplicons is performed to avoid competition between the reactions qpcr says for preamp are carefully optimized and validated by comparing analysis via priam and direct amplification using a standard sample deviations from expected Delta CQ with and without pre-emptive occasion reflect bias and standard deviations of replicates reflect reproduce abilities in this particular case only wanna say show poor reproducibility fewer say assays exhibit bias but that can be handled remaining assays perform very well the very large number of single blacks qpc ours required is performed in Nunnally two volumes using one of the two high-throughput platforms available to us the quan studio from life tech which is good for three thousand and seventy two independent qPCR reactions or the bio mark from phu ni where 96 samples are profile for 96 markers in a single run to handle and mine the large amounts of qPCR data generated we use the genic software and to correlate the results to biological functions we do pathway analysis using the I report so let’s see how this works in practice we were interested to study the response of ostracize to brain trauma using a mouse model that expresses green fluorescent protein under the control of the astrocyte marker GF ap we could sacrifice mice at different time points after trauma and collect single astrocytes using facts inspecting the measured expression profiles we recognize the log normal distributions but wait a minute this is not a normal distribution this gene which is vimentin shows two distinct Peaks evidencing heterogeneity most likely due to the presence of two subpopulations of austere science the measured data can be analyzed by traditional means comparing expression of genes one at a time before and after trauma using for example t-test calculating differential expression and collecting the information in Vulcano plot however this is not particularly powerful approach it suffers for multiple testing ambiguity as many of the apparently differentially expressed genes being false positives and it does not take advantage of genes correlated expressions genes are not expressed independently of each other rather groups of genes involved in the same expression pathway or part of the same network or expressing concert this can be exploited using multivariate methods multivariate methods classify samples based on genes correlated expressions most powerful is principal component analysis PCA the cells are shown in a scatter plot with the axes pc1 & pc2 that’s the principal components they are linear combinations of the genes and can be thought of being super markers virtual genes that have the combined properties of the real genes such that they best separate different types of astrocytes based on their expression patterns astrocytes from healthy brains are shown in blue astrocytes collected three days after trauma or shown in yellow astrocytes collected seven days after trauma shown in green and astrocytes collected 14 days after

trauma or shown in red in the graph it is seen that cells gradually move from top corner through the center towards the right side evidencing changes in expression in response to the trauma the astrocytes are becoming reactive even better separation is obtained in a three-dimensional plot which accounts for even more of the variation this is evident when viewing the 3d data using an interactive tool such as Gen X I’m now starting something known as the dynamic PCA in Gen X showing you the same data as in my slide the advantage is that here I can rotate and view the data from different angles to better see the separation then many of the genes that we have studied or included in our profiling experiment will not be sensitive to the conditions that we study these genes will not contribute to separation rather they will contribute noise by removing the non responsive genes from the analysis we shall improve the separation this can be done using a filter that removes genes with least variation across samples the filter is applied using the slider here so now I’m sequentially removing genes from the analysis and doing that I actually obtain better separation having done that before I know I will get a pretty nice result was about 18 genes remaining you can see now that there is a very very clear separation between the reactive astrocytes at 14 days after trauma shown in red and those corrected in the beginning you can also see that even here can separate the yellow green and blue but furthermore it becomes even clearer when I connect me lying astrocytes the reactive astrocytes are not a homogeneous population but rather or two sub clusters here evidencing there there are two groups of reactive astrocytes returning to the original two-dimensional PCA graph we can indicate the genes that are activated or suppressed as function of time after trauma as the suicides are activated we can also calculate the correlation between the different genes expressions among the individual sounds the correlation coefficient is a number between minus 1 and 1 and is calculated for each pair of genes numbers close to 1 indicated in green color are gene pairs expressed at the same time in the same cell numbers around 0 indicate genes that are expressed independently of each other a negative correlation which is not observed here would it indicate genes was opposite regulation when one gene is upregulated the other is down regulated correlation may be direct or indirect consider 3 genes there may be a master gene that induces the other two the dependence may be sequential or all 3 genes may be directly dependent on each other the difference between direct and indirect dependences can be illustrated was a more tangible example few years ago a Swedish newspaper studied the consequences of hot hot summer weather they found when the temperature was high Swedes were eating more ice cream that he also who found when the weather was hot drowning accidents were more frequent so they concluded consuming ice cream increases the risk of drowning sure they observed significant

correlation between eating ice cream and drowning accidents but fail to realize the phenomena had a common trigger that induces indirect or accidental correlation we can distinguish direct and indirect correlations by evaluating partial correlations using this strategy we can calculate expression networks indicating genes expected to be dependent of each other network analysis is a rapidly growing field offering important insight into biological functions classical network analysis is based on correlated expression of genes embolic that respond to common stimuli here we base the networks on genes expressed the same time but also in the same cell typically our networks are more distinct because of the much lower complexity of the responses of individual cells the genes we identify as being differentially expressed and relevant for the phenomenon studied came from Gen X be launched we have the internet for cloud-based pathway analysis using the ingenuity I report so now I’m on the Internet you and have logged into the ingenuity I report so here are my differentially expressed genes the sizes indicate for change between reactive and normal astrocytes and here for example I can identify the pathways the genes are involved in 10 of the genes are involved in the growth made receptor signaling which will soon be shown to me hopefully and this shows are also other genes involved in the same pathway that I can now include in further studies of expression profiling of my cells can you go beyond the single cell indeed we can this is a single cell or byte a very large single cell it is a noise site from the Frog sonatas Levis it’s not only large it’s two hemispheres have different shading this makes it easy to embed orient and mount the oil side in a cryostat and slice it we can then measure the transcripts in the slices which will reflect any intra solar radiance this graph shows intra cellar gradients of transcripts from the animal pole at the left to the vegetal pole at right we find these transcripts are off from the centre being closer to the animal pole other transcripts their most abundant in slices closer to the vegetal pole and most interestingly we find a third group of transcripts exclusively in the extreme slices at the vegetal pole these transcripts should be associated with a cell war we validated these results using the independent technique of digital PCR in digital PCR a single sample is analyzed using a platform having a very large number of reaction containers the number of containers should be of the same order as the number of target molecules in the sample when the sample is distributed into the platform most reaction containers remain empty while some contain a single molecule PCR amplification produces products only in those containers that initially contained target molecules counting the number of positive piece yards we know the initial number of target molecules that were present in the sample these are digital PCR data on the FDR side distribution of the two

transcripts are shown however now the oil side is only divided into five segments as you can see most of the acht 60 transcripts are found in the second and third segment on the oil side red color indicating positive peace yards while almost all went 11 transcripts are in the segment closest to the vegetal pole clearly the digital PCR and qpcr data are consistent the interest of the gradients of transcripts have consequences on the symmetry of cell division the first cell division of the oil side is along the animal vegetal axis and is not affected by the gradients we observe second cell division is also along the animal visual axis and is not affected by the gradients however the third cell division intersects the animal vegetal gradients introducing a symmetry among the eight customers formed notably however although the asymmetry is not manifested on the blasphemer level until the eighth sound stage it was present already in the beginning in the oil side these are some beautiful data collected on the bio mark using the 96 by 96 chip it’s a total of nine thousand two hundred and sixteen parallel reactions nice but so what we have seen many similar data before well you see these data reflect protein levels using the proximity extension assays developed at Ola link Biosciences we measure proteins the SE is based on pair of antibodies targeting the same protein the antibodies are tagged with oligonucleotides there are complementary in the three prime ends when and only when they are brought into proximity by the binding to the protein the three prime ends hybridize and in the presence of the polymerase and primers they initiate PCR the amount of PCR product formed is proportional to the initial amount of target protein that was present only one my career of crude sample is sufficient for analysis the workflow is based on pre amplification followed by high throughput qpcr and as before the data are analyzed using multivariate methods using genex this spring an oncology panel was launched which in summer will be followed by a cardiovascular panel and later in autumn there would be an inflammation path a related technology proximity ligation ASAE’s is available from Life Technologies here the oligonucleotide tethers have opposite polarity they are sealed by a complementary oligonucleotide and ligated by polymerase the ligated strand then serves as template for pcr also this reaction requires proximity and the amount of PCR product reflects the amount of protein target that was present we have developed protocol to lie as a single cell split the volume into three ala quads there are assayed for the presence of DNA RNA and proteins respectively cells from a sarcoma cell line were transfected was a vector expressing the Fu’s oncogenes tagged was green fluorescent protein with qpcr the amount of plasmid was measured with rtq PCR we measured Fu’s GFP mRNA but also related micro RNA and long non-coding RNA and was PLA PCR we measured Fu’s GFP protein

we see positive correlation between DNA mRNA and protein levels of force at the single cell level consistent with a central dogma of Jim Watson and Francis Crick we also found positive correlation between Mercer t1 cyclin d1 and small nucleolar RNA CD box 48 and negative correlation between the levels of ectopic Fu’s GFP DNA and cyclin d1 this is the first time we observe negative correlation on the single cell level finally let me just show some of the products we have developed at Terra for single cell profiling we have this Eliezer reagent that I mentioned for single cell lysis which is compatible with downstream rt-pcr they’re highly optimized reagents for reverse transcription pre amplification and PCR we have reagents for quality control the valid prime to measure genomic DNA background and compensate for it exogenous controls DNA and RNA spikes to estimate yields and test for inhibition we have even used them to micro inject them into cells and validating the entire protocol and intubate calibrators to remove variation between qpcr ons for analysis we use gen x to mine qPCR data and the I report for pathway analysis the services the offertory clients are in collaboration with leading companies such as wash Freud on life for mRNA expression execute and tray for micro RNA Holling for protein life tech for digital PCR thermo Fisher catch and measuring I go for sample extraction so thank you for your attention and I’ll be happy to take your questions okay professor kubista thank you very much for that presentation and at this time if you haven’t already done so feel free to enter your questions into the questions box and that’s located at the right hand side of your screen and the GoToWebinar software and yep you can click on the plus sign or the little out arrow in on the Mac and just pop that window out and type into it I’m going to briefly while people take an opportunity to do that I’m gonna just show you guys a couple of things that I DT has to offer so with that said for qpcr IDT offers our primetime qPCR probes and assays and I just I’m just putting this up here with some of the key features we have validated qpcr assays for human mouse and rat sequences and there’s there’s easy selection tools for finding your gene and for locating where it recognizes and such you can get five-prime nuclease based probe assays or you can get intercalating dye based assays that are primer only and then also since we’re talking about small volume PTR’s and things like digital PCR i just wanted to mention that we have a really great product called the zen double quenched probe and that’s it’s got two quenchers on the molecule and you get really low background you get really great signal to the background so these are very important things when you’re working with these small volumes we also provide a ton of free information as far as qpcr assay design we have our side tools design tools which from our homepage you can see we have this side tools drop down menu and you can find a list of all the things here and there’s various calculators in there so we have our primetime qpcr assays selection can be done from the side tools as well as you can do completely custom designs for PCR qpcr we have our logo analyzer that will do sequence analysis for DNA Allah goes and you your TMS and whatnot based on your reaction condition so there’s all kinds of free calculators the dilution and resuspension calculators for for whatever application and then we have a great deal of educational material we have a IDP decoded newsletter we write articles that are basically we offer core concepts and tips on various

scientific techniques including qPCR and synthetic biology next-generation sequencing RNA I a whole variety of applications and we do interviews with prominent researchers and talk about what they’re what they’re currently working on and what they’ve recently published so it’s just a really fun thing that we produce it’s great read it’s free and that’s at WWI DT DNA calm for its laughs decoded and then I also wanted to just bring up the bring up the qPCR application guide this is a free application guide that we have we talked about everything from basic qPCR types technologies instruments reactions setup troubleshooting there’s all kinds of great stuff in here you can get that free again from that wwii DT DN a comm and ford slash primetime so if you go on our primetime product page find the link for the free application guide and if you want any help finding answers to a specific question you can email our customer care at cust care at IDT DNA comm so having spent enough time on that I’m now going to move back to Professor Kavita and we will answer some of your questions so let me hand control back to him okay so we do have a you questions that are waiting for us professor Kapusta I think I have let’s see here sorry I had you muted there okay so the first question that we have is regarding micro RNA is it is it possible to do the single cell profiling in for micro RNA and is there anything special as far as the pre amplification step I don’t think in small-scale yes I mean we did measure micro RNA is in single cells but we have we are not preamplifier them because we don’t really feel comfortable about that yet however there are papers published on high-throughput mic or innate profiling that uses pre amplification however in our hands not yet okay here’s another question this is you’re referring to I think beta cells and this person wants to know if when you’re counting the transcripts per cell how do you ensure that what you’ve analyzed is a single cell and not a doublet or a cluster of cells that’s that that’s a good that’s a good question and it can be tricky in fact we very much rely on the fact sorting and initially it took quite some time to get a high success rate on sorting single cells it is possible to test that you get single cells because one can do for example a DNA analysis and you do also get a feeling of the total expression from the genes that you measure but that’s that’s essentially how you do it of course if you pick individual cells using a pipette then you have everything under control which actually if some of the dealer cell work was was not done by fax but was actually done using pipettes to pick individual of my cross creation on individual cells okay that kind of leads to this next question this actually came up earlier too is how can you normalize your expression for single cell experiments and do you use any reference genes it’s a very common question that’s why we had it earlier also it’s very important you can’t normalize single cell expression with reference genes because of the underlying burst kinetics because different different genes are expressed at different time points and different sounds however that’s not a problem because the most intuitive normalization is per cell so that’s what everybody is using in the field you just express the number of transcripts per cell okay here’s an interesting question so does does what you’re saying about single cell analysis does that also apply if you have a cloned expanded cell

population do you mean yeah whether you get a variability in our experience is that any cells that have active transcription shows this kind of variability because it’s a property of the mechanism of how genes are being expressed in cells if there’s dynamics we do see the variability in fact the only exception that I am aware of that’s actually all sides before they act initiate transcriptional activity so in the silent oil side you don’t have any variation then it’s actually perfectly read and they’re the perfect agreement between different oil signs ah very interesting this is a question that did not come up so can this technology be applied to plants well I believe so but it will be tough to lyse the cells you won’t have to have tools to lyse cells and analyzing plant cells is it’s much more trickier however probably one will have to do more do or introduce washing steps so there may be some losses but I’d certainly believe it can be standardized so the product call can be reasonably reproducible be aware that the differences between individual cells that we see are actually very very large which means that we can actually live with some conforming variation sure and as long as running that track I will ask this question – so how about fixed cells well they mentioned laser capture oh yeah laser capture cells laser capture can be analyzed fake cells usually the quality is really really poor and typically people are not analyzing fake cells so a lot of cells are fixed or for example using fax sorting and usually people do not try to analyze expression profile in those cells because the rest also poor so recommendation is avoid fixation of cells okay so given what we’ve talked about about reference genes what would you recommend for determining absolute transcript number for single cells and comparing the relative expression levels of your target genes well the termini absolute values is is well tricky or rather it does take some effort because one one really then has to calibrate for the reverse transcription reaction which actually which we have done in a few cases and we published some numbers on that typically the reverse transcription yield is between 0.5 and 80% so there’s a huge variation between transcripts if if the protocol includes a pre amplification step that also has to be calibrated for and then it’s also a matter of how to get the calibrator into the cell I mean we have done micro injection with our RNA spikes into the sounds but for most analysis that is relevant in in biology you don’t really need absolute numbers because typically you characterize the cells based on their profiles there’s a relative expression of the different genes and that does not require absolute numbers okay I think that this might be a bit of a general qPCR question so they’re asking is it sufficient to amplify approximately 100 base pairs of a gene motif for generalization of the regulation of that genes expression pattern and what they’re saying then they said in other words even if the PCR technology worked perfectly well couldn’t certain motifs be unpredictably masked and then impaired I’m gonna send this question you really quick let you read it yeah sure absolutely I understand everything that they’re getting at in that question it’s also possible to just send this person an email afterwards to clarify or if they can yeah yeah I think there’s more behind that question I mean there are there are certainly there are certainly

some sequences that are very very hard to amplify actually if I’m already some 15 years ago we we showed that some sequences could not be amplified in the presence of PCL buffers that contain potassium and then we actually also found the reason and the reason was that certain sequences formed intramolecular guanine quadruplex structures and they can’t be melted and in fact because of those reports from us many of the vendors removed potassium NTPC or buffers so that’s certainly one example and there may be others okay all right this next question yeah it’s a good one so regarding a single cell transcriptomic and now amplification and profiling do you have any experience looking at like single cell transcriptome and is there a method that you would prefer for that as far as quantitative analysis well if we were talking about amplifying the transcriptome that is amplifying the RNAs rather than doing the reverse transcription first we did look into that some seven years ago was the method that were available and in those days and they were not sufficient enough for single cell profiling in our hands these methods may have improved since then and because but we have not come back to that today we use the pre and PCR protocol okay sure so this next question comes back to the basics basic qpcr preparation again do you perform an RNA quality check on your single cell samples usually not because it would be too expensive in every sample we have to analyze some 100 cells but we have done it however beware the amount of RNA is very low which means that what cannot run classical capillary electrophoresis so however what one can do is to do differential analysis such that gene is analyzed using PCR reactions that amplify two fragments of different lengths of the same transcript and essentially by comparing the yields of these two amplicons we have an idea about the integrity of the RNA and in fact this works so well so we will be launching that as a product is autumn it’s turned out to be even more sensitive than capillary hearing to freezes and it is applicable to single cells but on a routine basis we don’t do that it would be too expensive okay this person is asking if accurate cell counts are hard to come by with facts for example it many events are a cellular you get you know some material it’s just not in a cell is it good enough to normalize the input RNA amount going into the are RT reaction it’s it’s it’s a it’s a tricky question because there were other aspects here it’s really not a good it’s it’s really not a particularly good ID tool to compare the expression in in single cells with two cells with three cells and so forth so the reason is that when the number of cells increases we are averaging out the dynamics of the births kinetics which mean that we can get we may see differences that are like different origin so to speak but but I’m not what the I think there are other problems I’m not really sure how you would even measure the total amount of RNA in a single Sal it’s just so little of it so it’s it’s it’s not as smooth method either I think that if there are issues in faxing the selves consider other methods to pick the cells laser microdissection can be used we have some experience of that and one can also use my cross for Asian to collect the cells and to to to capillaries and we have even used modified patch pipettes to actually open a sal and suck out the

cytoplasm that can be done even in tissues okay a little bit of a follow up on that so when you’re doing when you’re trying to like decide if your result is what you kind of expect for that cell do you usually look at multiple genes in each cell sample so you can like look and see all of my all the stuff I analyzed for the cell is you know twice as high as I would expect or something like that and does that kind of give you some sort of insight into whether you maybe have some sort of anomalous event yes yes it can it absolutely can we were not really certain about that before and the reason is that the amount of mRNA in a Cell also depends on the cell cycle however it’s it’s gene dependent so for example cycle you looking at the cycling genes I can tell where in the cell cycle cell is and and from that and then from the total amount of of transcripts one measures one could get a clue that this is probably not a single but this really we we don’t think that that’s a big problem I we are pretty confident that we are getting single cells and in most of these reactions and that that is fine I should also I could also mention that there is a new tool out there it’s the c1 from fluid I’m that actually allows you to collect individual cells and you can control that you have a single cell reaction ah interesting is that I’m assuming that probably captures them in a well set how that works well it actually captures it in a very small reaction chamber it’s a microfluidic system sure very cool I have to look into that okay oh let’s squeeze in one more question here this person wants to know how do you feel qpcr based techniques compared to nano string for single cell expression profiling are you familiar with that well it’s we don’t have any personal experience of singles our profiling with the Nano string technology from what I really works fine but one should be aware that also the Nano string workflow requires three amplification and I think that as long as you have to include the pre-application it does not really matter which technique you use if you major breakthrough would be if we could measure the transcripts directly without prehab that would be really great sure okay that’s about all the time we have there are still a couple more questions and we can look through them and see if there’s some stuff that needs to be responded to after the fact but you need to respect everybody’s time here so with that Professor kabisa I want to thank you again for this presentation we very much appreciate it it was very informative and I want to thank everybody else for attending today this webinar has been recorded and it’s one of a series of webinars at IDT we’ll be presenting on single-cell expression profiling and other topics we will email you about these topics as they’re scheduled and as a reminder you can find these videos recorded in there on our website at you’ll find them under the support tab in our video library and we’ll have our webinars and some other videos on our YouTube channel which is @ww youtube-dot-com forward-slash IDT DNA bio and with that again thank you everybody for attending we really appreciate it there were some great questions and yeah have a great day thanks professor kubista thank you