- Mikael Kubista added an answer:55How small of a "fold-change" in gene expression can be reliably measured by RT-qPCR?When using Sybr green or TaqMan qPCR assays to measure gene expression changes, can differences that are less than at least two-fold be taken as credible? What minimum fold-change or percent change is reliable?
Yes, that's indeed the case. When we report assay/test performance to clients we always specify separately LOD, LOQ in addition to classical parameters such s dynamic range and PCR efficiency.Following
- Yongzhong Zhao added an answer:1What are the statistics involved in dividing an expression ratio by another?
I am trying to reproduce a methodology described to identify changes in gene translation ratios using microarray experiments described in PubMed ID19804760, page 158. To calculate translational ratios, one divides the translation ratios from the experimental group by the translation ratio for the control group. My issue is that the translation ratio (Rn) in each group (control or experimental) comes from a microarray experiment that divides expression levels in ribosomal polysomes by the expression in ribosomal monosomes. Therefore each Rn (R1, R2, etc) has its own statistical significance. So for R1 it is p1 and R2 it is p2 (ANOVA analysis). How do we calculate the statistics of the ratio between R1 and R2 for example?
'Dividing an expression ratio by another' can be treated as a log-scale difference if with transformation. Once the GSE16738 data are log-transformed and normalized, it should be straightforward to conduct the analysis by using the 'limma' package with a 2-factor-2 level (3 repeats) design matrix.Following
- Lucie Grodecká added an answer:6How do I predict the function of a splice site mutation in silico?
If a mutation is in the splice site (donor), how can we bioinformatically predict its function in silico? For example if the site is broken or not, is there any exon skipping, if no then how to find the next splice site?
Many mutations of the splice site have been published, for
example, IVS1þ1G>A, IVS3þ5_6GA>AC, and IVS5-1G>C
in EVC, and IVS4þ2T>C in EVC2. Tompson et al. 
also reported mutation in EVC (IVS13þ5G>T) produced three alternative splicing in affected individuals.
How can we defend this mutation without expression studies, which is very difficult in Pakistan.
I agree with Florent. Recent publications conclude that one may use the in silico predictions as a first selection of mutations that possibly affect splicing. But they are still inaccurate even in determinig whether a variant will affect splicing or not - and they are not capable of predicting the splicing outcome (exon skipping, cryptic splice site utilization, intron retention). So the experimental validation is necessary.
If you want to try anyway, I prefer Sroogle engine (covering several programs for both the splice sites and the splicing regulatory elements predictions). In addition, there is a nice publication evaluating the outcome of splice site predictors: Houdayer et al., 2012. You might be interested.
All the best,
- Florence Mia added an answer:5How to detect histone methylation forms?
I'm trying to measure the activity of a methyltransferase enzyme for histones and would like to be able to quantify the results. I have tried using 3H labeled SAM and measuring the activity by scintillation counting, but the counts are usually very low and only 2-3 fold above background. I could potentially use a film and leave it for several days to increase the signal, but would like to try a faster method or a non-radioactive one. We need to be able to collect signal from mono- di- and trimethylatiion and don't have the antibodies available for all forms to do wb. Also,the amount of substrate that I have is very limited.What is the best way to do this?
A bit of clarification here: I'm doing an in vitro enzyme assay, so I have purified the methyltransferase I'm working with and have the histone substrates and I measure the activity by using 3H-labeled SAM and scintillation counting, but the counts are very low and doing wb is not an option either since antibodies are not available. What other method could I possibly use?Following
- Alexander Ruthenburg added an answer:6Did anyone also recognize a loss in efficiency over time of Zymo-Spin CHIP kit?
We use the CHIP and purification kit from ZYMO research for CHIP experiments. Now we even took double amount of cells (neurons) but get almost no chromatin anymore (<10ng/ul). The first experiment was fine, but each time we repeated it we got less DNA after shearing, reverse crosslinking and purification. Is there any possibility that we lose DNA somewhere? Or does multiple loading affect the efficency of the columns? Can the columns lose their quality somehow?
Thanks a lot!
Agree with Dan about sonication-- bath sonicators actually decline over time in energy delivered to sample (I suspect this is due to the sonication element being glued to the bath and this coupling erodes with use/time).
Also, we never use kits, but I would imagine that formaldehyde degrading in solution rapidly is a problem here. We always make it fresh from PFA. See other posts on fresh paraformaldehyde cracking.Following
- Iman Aldybiat added an answer:11Does anyone have experience with DAVID bioinformatic program?
I am doing the analysis of my data with David programme. The problem or what I get from the analysis is :similar results when I do interpretation of two conditions when these are up or down regulated. For example: the analysis of A vs B up regulated gives: phosphoprotein 35%, P value 1.2 E10-7. the analysis of A vs B down regulated gives also phosphoprotein 31%, P value 7.4E10-2. The same thing with other molecular functions (ion binding, metal ion binding, zinc finger....). Is this normal ? or shall I use another more specific parameter than percentage of involved genes? thanks a lot
Thank you Marco. My gene-expression data are from microarray. It is true that I did not looked at the background which is Mus musculus in my experiment. When I do analysis with DAVID: I do upload of my data and then I choose the identifier (i.e. Agilent oligo) and gene list before submit list. I am actually concerning on doing the analysis via IPA ingenuity. What was useful for me and avoided to see a lot of molecular functions (ion binding, cation binding,.....etc) is that I reduced the number of genes before being treated with DAVID.Following
- Tamás I Orbán added an answer:10MicroRNA binding sites at mRNA?
Many papers show that microRNA regulates gene expression at the post transcriptional level (If microRNA increases and its target gene decreases or else microRNA decreases and its target gene increase).
My question is whether microRNA and mRNA are both positive regulation. (If microRNA increase and its target gene increase or microRNA decrease and its target gene decrease).
Just a small addition: there was a paper showng that the same miRNA can also increase TRANSLATION from its target mRNA in a cell cycle dependent manner (Vasudevan, S., Tong, Y., Steitz, J.A., Science, 2007, 318:1931). In such cases, the problem is always to exclude the indirect effects - but in that article the results seemed convincing.Following
- Doinita Ispas added an answer:10Optimal primer concentration for real time PCR?Can anyone please tell working concentration of primers (in nanograms/picograms) for real time PCR experiment. Because I have primers stock at 2 microgram concentration.
--6126 g/mol means 6126 ug/umol
--if 6126 ug is 1 umol, then 198.4 ug is 198.4/6126= 0.0323 umol, i.e 32.3 nmol
--dissolving 32.3 nmol in 323 uL pure water is giving a stock solution of 100 uM (or 100 pmol/uL)
Use this stock or make a less concentrated intermediate working stock, to your convenience.Following
- Zhongyi Hu added an answer:5What is the appropriate svm type and kernal type and parameters (c,gamma, etc...) in LIBSVM for microarray data?
I am working on optimizing Gene selection in microarray data for Cancer Classification. I am going to use SVM in (libsvm) as wrapper approach to evaluate Gene subsets using 10 K fold cross validation.
Microarray data consider as huge dimensional data ( i.e Lymphoma data set consists 4026 Genes 'features' and 62 instances and 3 class labels).
Does libsvm support multiclass classification, As in my work, Lymphoma & MLL has 3 classes?
What is the appropriate svm type and kernal type and parameters for the chosen kernal (c,gamma, etc...) in LIBSVM multi class classification like microarray data?
two papers should be refered to when you employ Libsvm:
LIBSVM: A Library for Support Vector Machines
A Practical Guide to Support Vector Classification.
You can google it to find the pdfs.
the famous and simple way for parameter optimization of svm is grid search, please refer to:
A Practical Guide to Support Vector Classification.
Some of my studies use meta-heuristic methods to optimize the parameters, please read the following paper for a brief overview about parameter optimization of svm:
A PSO and pattern search based memetic algorithm for SVMs parameters optimization http://www.sciencedirect.com/science/article/pii/S0925231213002038Following
- Nicholas Wagner added an answer:4Can someone advise me on DNA methylation studies in animal (rat) tissues?
I am doing bisulfite conversion experiment and I don't know how to design my experiment. I use rats from different ages to see if there is a difference of the methylation status of one promoter during development.
But I don't know how to organize my groups of animals. In literature, this part is poorly developed so I don't know if I can look at the methylation status of pool(s) of animals or animals individually. And is one pool enough to make a comparison on the methylation status between 2 ages?
The first experiment I did was one pool of 6 young rats and one pool of 4 old rats, but I don't know if this is correct, if I have to replicate with other pools or if I have to assess the methylation status of each animal individually.
I think this method is more qualitative than quantitative.
Thank you for your answer!! Have a nice day :)
Question 1: I don't know how to organize my groups of animals. In literature, this part is poorly developed so I don't know if I can look at the methylation status of pool(s) of animals or animals individually.
I know that there are very well written publications in this field - look for the ones that are well-written (and be more careful criticizing the publications of people in your field...)! I am quite sure that details are well-described well in the publication I was part of attached below (Kovacheva et al.). If you have any more questions please message me directly here on RG I will be glad to help.
However, as a quick answer for all, usually each animal is measuerd individually, and then the results are combined/pooled afterwards if necessary, as already stated by Leopold Fröhlich. This of course means (and I would seriously recommend) doing the bisulfite treatment individually for each DNA. After this you have the option of pooling for measurment (which I would not recommend, unless inevitable), or measuring and then pooling results. The latter is the more common and what we do in our lab. There are, of course, exceptions: if material is limited, for example when investigating oocytes, then you may be forced to pool before bisulfite treatment or measurement to have enough material - but this should not be the case for rat tissues, which are comparatively large.
Question 2: The first experiment I did was one pool of 6 young rats and one pool of 4 old rats, but I don't know if this is correct, if I have to replicate with other pools or if I have to assess the methylation status of each animal individually.
For a first-off check I guess this would have been ok. It would have been better if you had measured animals individually, as pooling may obscure small differences. The consequence that I think this has for you is this: if you see differences between young and old in your analysis as it is now, it makes sense to do a more strictly planned second analysis. However, if you do not see differences, it may be because pooling is hiding small differences.
In general I would say there are better ways to approach a longitudinal study/ study over time. For example, you could draw blood from the same animals at different ages, as Somnath suggested. That way you can follow the methylation of each animal over time. This will not be possible if you need to sacrifice the animals for your analyses. Then it would be best to compare siblings from the same litter (or at the very least form the same mating pair), and sacrifice some at young and some at old age. You will need setups like these as otherwise other factors, such as the differences between animals of one strain, may influence/obscure your results too heavily and hide the methylation increase over time that you are looking for.
Question 3: I think this method is more qualitative than quantitative.
There are qualitative and quantiative ways to measure bisulfite-treated DNA. You do not state which method you are using after bisulfite conversion. I guess it is Sanger sequencing? If this is the case you will need to sequence mulitple clones and quantify based on that - it will not be possible to quanitfy based on the peaks of only electropherogram! Could you please go a bit more into detail? Then I can tell you more :-)
Hope this helps and good luck with your analyses!
- Yair Botbol added an answer:5Why do I have upregulated expression in PCR and microarray but I see downregulation in Western blot test?
I worked about mRNA expression on RCC. One primer shows upregulaton in PCR and microarray but express downregulation in western blot.
2 possibilities appear to me so far:
1-post transcriptional: possibility of non-coding mRNA but I am not enough familiar with this field so I cannot give you suggestions.
2-post translational: to test this hypothesis I recommend to start with the exp I suggested above (protein degradation test)Following
- Radoslaw K Ejsmont added an answer:3What is the best way to identify known transcriptional regulators of a Drosophila gene set?
e.g. 299 genes have altered mRNA expression in our tissue/mutant/etc. of interest. I want to find out all the transcriptional regulators of those genes. We're working in Drosophila melanogaster.
You can try i-cisTarget - it's a tool to identify overrepresented regulatory motifs (and thus known TFs) in the specified gene set.Following
- Bharath Reddy added an answer:4Can I calculate heritability for augmented design with 1 rep & 1 loc for 1 repeated and 2 random checks?
In 2014, I had planted 210 lines, 3 checks (1 repeated check & 2 random checks) in augmented design 1 replication, 1 location ( 2 loc were planted but lost 1 loc for late freeze damage). These 210 lines comes from 12 populations (Family structure is complex I have 7 wild relatives back crossed to 2 elite parents). In 2015 out of 210 lines, 93 were advanced to next generation based on tillering ability (alpha lattice, 2 replications, 4 locations). I have done BLUPs and BLUEs for 2015 using META & I got heritability for 2015. I have done moving mean analysis using Agro Base Gen II for 2014. End of the day I have to calculate genetic gain we have achieved for grain yield by indirect selection for tillering ability. Please guide me step by step
Thank you sir I will talk to Rupa & other statisticians. If I can't answer I will get back to you.
- Lekha Dinesh kumar added an answer:4Can shRNA or SiRNA knockdown a highly expressed gene as efficiently as genes that are expressed at a lower level?
This might be a naive question, but I just wonder if there are more transcripts in the cells, would it be harder for the knockdown mechanism to degrade all these transcripts? Imaging for a same target gene, which expresses at a very high level in cell type A but at a lower level in cell type B, if you use a same shRNA or siRNA, would you expect differences in knockdown efficiency?
I feel it will be more efficient to shut down a higly expressed gene since the siRNA signals will be quite amplified,thus resulting in an efficient knock down!Following
- Lesya Holets added an answer:34Troubles with RNA extraction from mouse skin.I'm trying to obtain RNA from mouse skin but the results have not been satisfatory. I observed chemical contamination and also very low amounts of RNA (1-5ng). I'm working with a murine model who has collagen overproduction and usually freeze the samples directly in -80°C in RNAlater.
Qiagen support told me that the RLT buffer can crystallize and that I should use a water bath at 37°C after this specific step to dissolve the crystals (this is not recommended in the kit). Do you think this is possible even working at room temperature? Any other recommendations ?
I keep the tissue in RNAlater at 4C for 1-2 weeks, about 1 mo at -20C, or -80C for longer time period.. I used GeneElute mammalian total RNA miniprep kit from Sigma for RNA extraction.Following
- Jochen Wilhelm added an answer:9Is it best to present your qPCR data as a ratio or log(base2) (aka, ΔΔCt)?
There seems to be a large variation in the presentation (and analysis) of qPCR data. I am wondering if it would be best to present my data as a ratio (2^-ΔΔCt) or take the log(base 2) approach (which just converts the value back to ΔΔCt)?
Assuming Efficiency =2 :
ΔCt(treatment) = Ct.target - Ct.reference
ΔCt(untreated) = Ct.target - Ct.reference
ΔΔCt = ΔCt(treatment) - ΔCt(untreated) (from Livak & Schmittgen, 2001)
Additionally, as far as I have understood the literature and previous posts here, it is necessary to do statistical analysis (ANOVA, t-test, etc.) on the ΔΔCt value, not the ratio, correct?
Thank you very much for your input and help!
Yes, it's actually quite simple :)
Thank you for your feedbak. I am happy that I could help.Following
- Tomas Pereira added an answer:8Why is my qPCR threshold line is below the background?
We are having trouble to determine the threshold line of qPCR in a ABI 7500. We always used automatic settings for determination of the threshold line, but recently the software is setting the threshold line below reaction background. Do you have any idea about what is causing this problem and how we can fix it? Attached is following an amplification plot of the wrong threshold. Thank you.
I would like to thank you all for your answers and for trying to helping solving our problem.
We recalibrate our system using a new Spectral Calibration Kit as recommended by Laurent and other researchers, and we performed a new qPCR and the same problem occurred. After that, we contacted Life Techlonogies support and I explained our problem. We discovered that the real problem actually was in the FAM probe. Life Tech is actually providing a new probe for us.
We think that a new probe, our problem will be solved. Thank you Kevin for your answer and help.Following
- Jason W Hoskins added an answer:6What are the possible mechanisms through which SNPs in the promoter region of a gene could affect clinical outcomes ?
SNPs in the promoter region of a gene would be expected to modify the binding affinity of the promoter to transcription factors thereby regulating gene expression. Are there alternative explanations for evidence linking promoter SNPs to diseases ?
On another note, if this promoter SNP is not linked to the risk of onset of a disease but to adverse clinical outcomes such as mortality or readmission rate, what explanations could be provided to explain such an association ? Could it represent a role in early versus late phases of the disease ?
A few notes:
First, the functional SNP most likely mediates altered protein binding, though other possible mechanisms could include altered CpG site affecting methylation of the region, or altered ncRNA binding. However, the latter 2 possibilities are more theoretical since the causality for the association between DNA methylation and expression is still murky, and I'm unaware of any study showing a functional SNP altering a ncRNA binding in a promoter. The effects of altered protein binding vary greatly depending on the protein, but could include loss/gain of enhancer/repressor TF binding, nucleosome repositioning and altered long-range chromatin interactions. There are numerous methods for determining allele-specific protein binding, including in silico binding prediction, supershift EMSA and mass spec of oligo pulldowns.
Second, even when you have identified the actual functional variant (which is not simple), determining the gene or genes affected by the variant is no trivial matter, even if the variant lies in or near a gene's promoter. As the downstream biological implications of the functional variant will stem from the gene(s) affected, this needs to be established. I have seen cases where a functional variant is very close to the promoter of a gene, but rather than affect that gene, it alters expression of a more distant gene. Tools like eQTL analysis, 4C-seq and CRISPR/Cas9 genome editing are great for clarifying this.
Third, in regards to your questions about the SNPs effect on clinical outcomes, I do not think there is any way to even begin to answer such questions without establishing the gene(s) that is/are affected. The mechanisms for how these SNPs might effect clinical outcomes are as numerous as the effects of genes and gene networks. I don't think there is any shortcut to get from SNP to disease measure mechanism without identifying the gene(s).Following
- Leavy Zhang added an answer:4Is there any method to classify epigenetic peaks according to peak shape?
Recently, I was doing Histone modification analysis using ChIP-seq. I found that different peak shapes occurred for one specific modification (i.e. H3K4me3). So, I want to do a clustering according to their peak shape? Perhaps simple methods like K-means or hierarchical clustering might not be the very choices for this.
Could someone giving me any related advices? Like signal recognition or model-based clustering method?
Thank you, Anil Panigrahi! I do agree with your comment on factors impacting histone modification peak shapes. However, it's a great pity that my data sets have no biological or technical replicates. So, it's difficult for me to investigate peak shapes across experimental or biological replicates. What I am always concerning about is that what these peak shapes are related to gene functions and competitions between different histone modifications, and this is the very story that I want to see with my scanty data sources.Following
- Hemant Prajapati added an answer:5How can I do semi-quantitative measurements for ChIP assay?
I am performing Chromatin Immunoprecipitation assays (ChIP) and wanted to do semiquantitative measurements for my different ChIP experiments. Can anyone suggest me the simple and reliable method for this. Also which software can be used if I have saved all my gel images into .tif format. If I take all conditions same for WT and mutant then is it necessary to include WCE (Input) into calculation?
Thank you all for useful information.Following
- Pranay Amruth Maroju added an answer:3What is the maximum number of missing loci allowed to exclude an individual from an analysis (e.g. Geneland, Structure)?
I'm trying to figure it out what is the maximum number of missing loci allowed to exclude an individual from an genetic analysis (e.g. Geneland, Structure)?
I'll be grateful for any help
It depends on total number of loci you have, population size and relatedness of the individuals for microsatellite data. CERVUS software will tell u whether you have enough loci to perform genetic analysis in the form of a line graph (Probability of identity on Y axis Vs No. of loci on X axis). Also if your population size is less hardly you can allow any missing locus. More related individuals require more scored loci.Following
- David A Armstrong added an answer:2Has anyone used the spike-in for circulating microRNA?I would like to know if there is anyone who has worked specifically with QIAGEN cel-miR 39 spike-in using the QIAGEN miRneasy mini kit.
I have not used this specific spike in , but keep in mind that spike-in's are only a control for sample processing, not for biological variation... there has been no widely established normalizer for circulating microRNAs yet - best choice so far geometric mean.Following
- M Sadman Sakib added an answer:5How does miRNA regulate protein fold?
Animal microRNAs (miRNAs) regulate gene expression by inhibiting translation and/or by inducing degradation of target messenger RNAs. As we know, in addition to downregulating mRNA levels, miRNAs also directly repress translation of hundreds of genes. MiRNA can, by direct or indirect effects, tune protein synthesis from thousands of genes. But it is unknown how can miRNAs regulate protein fold.
I was also having the same thought. But didn't find any good papers regarding this. I think this could a an emerging field of research.Following
- Tomasz Jurkowski added an answer:1Why do methylation changes occur so quickly as compared to transcriptional reprogramming?
Ngn3,Mafa,Pdx1 mediated lineage conversion ,converts acinar cell to beta cells .Analysis found that DNA methylation changes in 10 days and acinar cells appear to be insulin + but transcriptional reprogramming takes large time ,,,about 2 month
Can any body help me to find out why ?Following
- Tommaso Andreani added an answer:6Does anybody here have experience indexing vcf files using tabix and vcf-sort?
I am trying to use vcf-merge on 2 of my vcf files in order to carry out an Fst analysis in the software; for that I need to use tabix and vcf-sort to gunzip, sort, and index my file.
Ive successfully gunzipped and sorted the files. I am just now having troubles indexing them, because tabix returns the "Chromosome blocks at [position] are not continuous. Has the file been sorted?"
Any help would be appreciated from any bioinformatics superstars!
Next time you do not need to gunzipp 'cause it takes more time.. Happy you have solved!!
Cheers and vote!Following
- Anna Git added an answer:6When is it useful to apply the 40-DeltaCt method for calculating relative gene expression?
I have recently come across a clinical study that expressed gene expression in the following way: "RNA results were then reported as 40-DeltaCt values, which would correlate proportionally to the mRNA expression level of the target gene." (Where delta Ct was the difference between the Ct values of the gene of interest and a reference gene. In this case 40 cycles were used for amplification.) In what type of experiments is it useful to apply this (40 - delta Ct) calculation? How does this relate to the more frequently applied 2(deltaCt) - method?
I completely agree with the emptiness of "40-Ct" transformation. Some of my experiments only run for 30 cycles (abundant snRNAs) or 45 cycles (pre-amp + amp of rare miRNAs).
In general, I'd be very very cautious in applying ANY of the above Ct-based calculations, as they assume that the amplification of both genes (reference and interest) is of equal efficiency. This is rarely the case, and over a large number of cycles, differences creep up. We include a standard curve for each gene and convert Ct values to relative input, which can then be normalised in any way you see fit. Also include a titration of RNA input into the reverse transcription. You'd be surprised how many commercial kits are saturating their reactions, which affects abundant and rare transcripts in different ways!
We also stopped using a single reference gene, but use a geometric mean of 3-4. This cannot easily be done with ddct, unless you accept averaging Ct as geo-averaging input etc. When dealing with a perturbation experiment in cell lines, it does not matter. But when dealing with heterogeneous samples (e.g. patient material), it is a much more robust measure.
Lastly, I would urge everyone to consider the MiQE guidelines championed by S.A. Bustin.
About Quantitative Gene Regulation
A group for scientists interested in quantitative descriptions of gene regulation in pro- and eukaryotes (equilibrium and non-equilibrium protein binding, chromatin rearrangements, covalent modifications, input-output cis-regulatory functions, etc) using approaches of biophysics, molecular and cell biology, bioinformatics, systems biology and synthetic biology.