Science topic

Cancer Genomics - Science topic

Explore the latest questions and answers in Cancer Genomics, and find Cancer Genomics experts.
Questions related to Cancer Genomics
  • asked a question related to Cancer Genomics
Question
1 answer
As in cancer genomic studies, we have databases such as TCGA and CBio Portal, do we have a similar database for hematological disease? These are databases where we can explore the hemoglobin gene but I am specifically looking for the whole-genome database. A dataset of WGS of a couple of patients could also be very helpful.
Thank you
Relevant answer
Answer
One of the oldest locus specific databases for hemoglobin disorders is HbVar available at http://globin.bx.psu.edu/hbvar
You can also check for Hemoglobin variants in the Biorad Library
You can also check hemoglobin variants on Ithanet
  • asked a question related to Cancer Genomics
Question
8 answers
Standard plasmids degrade, but episomes do not. Why?
Relevant answer
Answer
An episome is an integrative plasmid. Episomes replicate together with the rest of the genome and subsequently associate with metaphase chromosomes during mitosis. The integration into the genome allows stable maintenance of the episomal DNA over several generations. As an example, DNA in some viruses such as herpesviruses, adenoviruses, and polyomaviruses serve as episomes.
Examples of episomes include insertion sequences and transposons. Viruses are an ideal example of an episome. Viruses that integrate their genetic material into the host chromosome enable the viral nucleic acid to be produced along with the host genetic material in a non-destructive manner.
Another example of an episome is the F factor. The F factor determines whether genetic material in the chromosome of one organism is transferred into another organism.
Best.
  • asked a question related to Cancer Genomics
Question
3 answers
CRISPR/Cas9
-can eliminate DNA doublestrand breaks(DSB)
-enables hope curing chromosomal abnormalities and induced maladies
However, chromothripsis should be taken into serious account to remain objective in judging it
#crispr #crisprcas9 #Chromothripsis #mutation #genomics #micronucleimodel #science #research
Relevant answer
Answer
Could be
Can be
Might be
....
However, it is up to scientists judging and valueing facts (not opinions, hope, ..)
Thanks
Björn
  • asked a question related to Cancer Genomics
Question
4 answers
I am working on cancer genomics. I downloaded TCGA BRCA FPKM data for my analysis. after some preliminary analysis, I categorised 250 samples in total 2 distinct categories. Now I want to analyse and compare their epigenetic and mutation patterns.
After downloading all mutation and epigenetic data from TCGA BRCA, none of the samples from my previous analysis is matching.
Suppose, TCGA-AN-A0AK-01A-21R-A00Z-07 sample is present in my previously downloaded FPKM data.
But TCGA-AN-A0AK-01A-21W-A019-09 is available in the mutation data. Are these ids the same? For FPKM data and mutation data, do the same sample (individual) be represented by different Ids?
Please help me, Thank you in advance.
Relevant answer
  • asked a question related to Cancer Genomics
Question
5 answers
This might be a very vague question, but I'm trying to understand what the common stumbling blocks are when trying to make full use of cBioPortal?
For example, researchers are limited to querying only 100 genes from the cBioPortal.
What sort of work can you do through R/python environment that you can't do directly through cBIoPortal?
Relevant answer
Answer
The limitations are imposed to prevent the server being overloaded with analyses but if you create your own instance with any custom API, there should/might be any limitations and analyses can be applied to a wider candidates. However, one must how knowledge and understanding of CLI operations and troubleshooting.
  • asked a question related to Cancer Genomics
Question
2 answers
The project's budget is 12,000$ does not include buying any equipments except (for example) a genotyping analysis kit, I did a project for analyzing genetic diversity and selection signatures in four endangered cattle breeds using Illumina BovineHD kit but it was not satisfying, any suggestions? it is very important and crucial for my career.
Thanks in advance,
Relevant answer
Answer
You can also choose to study the diversity, evolutionary phylogenetics or domestication of a species. Large stractural varitions on genetic disease is also a choice.
  • asked a question related to Cancer Genomics
Question
6 answers
Cbioportal provides an online tool for generating oncoprints but I want to generate one for customized data. It would be great to know if a tool is already existing for this. 
Relevant answer
Answer
Here is a Python implementation: https://github.com/pjb7687/pyoncoprint
Disclaimer: I am the author of the package.
  • asked a question related to Cancer Genomics
Question
4 answers
Has anyone have used TCGA analysis to get useful publications. If yes, would it be possible to get a copy of R code to use it as a road map for the procedure.
Relevant answer
Answer
  • asked a question related to Cancer Genomics
Question
1 answer
Gene copy number variation is one type of genomic diversity among different subjects.
Now, over time, people accumulate somatic mutations in their genome. Some of these may include gene duplication or deletion.
Has anyone addressed the question whether, in a given individual, the number of copies of a given gene can change over time in a particular cell lineage? Or in large cell populations (PBMNCs)?
Relevant answer
Answer
Not sure about CNVs, but more and more studies are focusing on somatic mutations in healthy tissues. Examples are:
  • asked a question related to Cancer Genomics
Question
1 answer
Dear Naema
I am a bioinformatics scientist focusing on cancer genomics. I recently came across you paper 'Genomic characterization of human brain metastases identifies drivers of metastatic lung adenocarcinoma ' and found very impressive.
 
      We are doing some research similar to yours on identifying significantly altered CNV events between two groups. I am wondering if you mind sharing your source code for identifing brain metastatic drivers corresponding the part of Copy-number driver analysis in the methods? This reference would be really helpful to our study.
Best,
Cheng
     
Relevant answer
Answer
I guess you should contact the editor of the journal if the authors do not answer back. Most likely you will not get an answer here.
  • asked a question related to Cancer Genomics
Question
5 answers
Can you combine and analyse RNAseq data from different TCGA datasets? I am asking specifically about STAD and ESCA, as STAD contains several gastro-esophageal adenocarcinomas which I would like to analyse together with the adenocarcinomas from ESCA to construct a larger 'esophageal adenocarcinoma' cohort.
I am wondering if each dataset has gone through a different pipeline meaning that it is not possible to do this (due to different batch effects etc.)?
Relevant answer
Answer
My advice would be to look for harmonized data (I guess it means all data from raw to count matrix processed in the same manner) and then adjust for batches, covariates - e.g. study IDs at the modelling level. Going into details, you may have to address statistical issues like exchangability and ignorability.
  • asked a question related to Cancer Genomics
Question
3 answers
Hi friends,
I need breast cancer gene sequence for indian women. where can i download the data?
Relevant answer
Answer
1. Identify which gene you are looking for
2. Sequences would be same as that of Hg38 build, take concesus including causal variants of breast cancer in Indian ethinicty (use: dbSNP/COSMIC/Clinvar).
  • asked a question related to Cancer Genomics
Question
2 answers
Hi
I am exploring the sORF-encoded peptides which are less than 100 AAs long. As i read these kind of peptides are the direct product of gene; they did not came from processing of large precursor proteins. So, i need to clarify my doubt whether the signal peptides are present in this type of peptides or not. Any one working or having the knowledge of these peptides please clear this doubt.
Thanks  
Relevant answer
Answer
I can suggest a few papers if you are interested.
  • asked a question related to Cancer Genomics
Question
4 answers
I would like to conduct a project on analyzing commonalities of certain mutations in repetitive DNA of patients with certain types of cancers. To do this, I will need to find the whole genome sequence of different patients with the same type of cancer. I would very much appreciate it if you could direct me to a database where I could obtain this information, or if you have any tips on where I could find this information!
Thanks from a grateful undergraduate,
Nitish
Relevant answer
Answer
Marisol Herrera-Rivero , Ali Javadmanesh , and Nuno A Fonseca : thank you so much! I appreciate your answers, and am currently looking into them. As I investigate further into the links each of you have provided, I may have further questions if you don't mind, but I truly appreciate your help!
  • asked a question related to Cancer Genomics
Question
7 answers
i want study cancer genome by bioinformatics tools. can advice me in articles or review paper can guide me to do my research.
Relevant answer
Answer
You will no doubt want to start by getting familiar with The Cancer Genome Atlas (TCGA) website and portal:
Here is the data portal: https://portal.gdc.cancer.gov/
And, this documentation provides useful guides to the bioinformatics pipelines used/available: https://docs.gdc.cancer.gov/Data/Introduction/
To begin reading you may find this review article helpful:
Best,
Chris
  • asked a question related to Cancer Genomics
Question
4 answers
By certain type of cancer, mutations are well known. Genome/epigenome editing could possibly "repair" such genes or silence them by methylation of their promoter.
Relevant answer
Answer
In my opinion, owing to the vastness of the genetic and epigenetic alterations in cancer the only way forward is to get rid of the mutated cells. Genome editing of immune cells could help achieve this but I don't think we will ever be able to revert the actual cancer cells back to their former 'healthy' state.
  • asked a question related to Cancer Genomics
Question
6 answers
Dear all,
I’m performing a cancer genomic analysis and I need a negative control of non-cancer genes (200-400 genes). So, I appreciate any help.
S.
Relevant answer
Answer
The first answer while technically correct is a bit misleading: COSMIC reports all somatic mutations found in cancer samples. This means that many of the genes touched by these mutations are not what people think when they speak of cancer genes: they generally want genes that are directly involved in driving a cancerous state (ie genes with driver mutations and not passenger mutations). So one solution is to substract from the complete list of human genes (on Ensembl, HGNC or NCBI) a master list of cancer genes. The problem is that there is not a single consensus list. COSMIC keeps a curated census list:
That contains 723 genes
But this is not complete and 2 years ago I tried to compile myself such a list and at that time I had already >3'000 genes in it.
Note that if you do go with the first answer your list will also be incorrect for the following reason: some genes are not 100% mapped correctly by Ensembl and thus Cosmic can not in some cases map somatic mutations to these genes. As there is a huge amount of data now in Cosmic almost all genes that are mapped have associated sequence variations, so the list you will obtain will contain a mixture of very small genes and incorrectly mapped genes.
Best
Amos
  • asked a question related to Cancer Genomics
Question
4 answers
Dear Researchers,
I am working with a PDCL cancer cells and I want to transfect GFP in the cells. I am using lipofectamine 2000 and I incubated it with the cells for 6 hour and seems to be good but after 6 hours they start to show some death due to Lipofectamine toxicity.
So I only incubated cells with the Lipofactamine and the vectors for 6 hours and then changed the medium to the normal medium that I am using for the cells.
I used in this experiment 10.000 cells and they showed some positive cells after 72 hour however they all died when I tranfered them to a larger T25 Flask,
Any suggestions ?
Relevant answer
Answer
Hi Sandra,
Thank you very much for your valuable input and looking forward to using Viromer in our cell line setups
Best Regards,
Mohammed.
  • asked a question related to Cancer Genomics
Question
2 answers
I have been looking into cancer The Cancer Genome Atlas with no luck. Where could I find studies comparing samples prior and post any kind of triple negative cancer treatment? Comparison of microscopy imaging, expression, etc.
Thank you
Relevant answer
Answer
Thank you!
  • asked a question related to Cancer Genomics
Question
1 answer
“Oncomine Research Premium Edition is designed for laboratories engaged in advanced cancer genomics research and can be purchased in 2-user or 10-user packages for $5,000 and $10,000 respectively. ”(http://www.bio-itworld.com/newsitems/2007/oct/18-briefs/)
So I almost have no chance to use this edition for lacking money?And the researchers won't borrow the account to strangers?
  • asked a question related to Cancer Genomics
Question
6 answers
What is the difference between RNA sequencing and Exome Capture RNA Sequencing and how it differentiated with exome and RNA Seq?
  • asked a question related to Cancer Genomics
Question
1 answer
I have been searching for SNPs in miRNA genes (formulated panel of 65) in a designated breast cancer population by iPLEX/MALDI-TOF-MS. But after that just got stuck into and couldn't go further.
Expression analysis was suggested by someone, but all I need is some strong, supportive strategy to back my genotype data.
All I need is a follow-up strategy.
thanks in advance.
NB: I am just at the very beginning of my PhD career on Cancer Genomics. So, need some expert opinion to walk through the path.
Relevant answer
SNPs which  occur at miRNA-related regions it may change in  phenotype.  still there is need to know about the  evolution of miRNAs after SNP variation, there is possibility to  perform  genome-wide scan of miRNA-related SNPs, need to analyze their effects on the stability of miRNAs structure and the alteration of target spectrum in plants  for more information refer following link.  
  • asked a question related to Cancer Genomics
Question
4 answers
It's beyond doubt tailor made treatment options for a specific diseases and patients will help the health care system better. Amongst other things individual therapy by gene IS a viable alternative. But for the economic aspects. 
Relevant answer
Answer
Do you make money more important than patient's life? Don't you care about the accumulated drugs' toxins in the environment and patient's prescription drugs are adding more and more?
Even bargain for lower drug's price. Eventually, in overall, you make more money.
  • asked a question related to Cancer Genomics
Question
5 answers
Please Specify if you would be interested in reading an article about this topic.
Relevant answer
Answer
"In one of my researches of the case report, amputation surgery spent a lot money and medical resources. But, all patients die within 5 years without a quality life plus metal suffering. Is it worth to do surgery? Besides, what surgery to do? The personalized care can put "should I do surgery or not? Is there a way to save the patient without surgery" as another out of the box thinking?"
Resources at hand versus the outcome (quality of life): In ancient native Indian tribes, it is claimed that the very old and feeble, with no realistic hope of improvement in their health, would say their goodbyes and simply go out of the warmth of their tepees to expose themselves to below freezing temperature for a relatively quick and peaceful end. It's the same old "box" but perhaps being relabeled. 
  • asked a question related to Cancer Genomics
Question
3 answers
In cancer cells, can autophagy inhibition leads genomic instability in order to promote cancer progression?? thank you for your help. please provide references if you will find
Relevant answer
Answer
Many thank for your kind reply.
Best regards 
  • asked a question related to Cancer Genomics
Question
3 answers
I have tried COSMIC, CCLE and the Achilles Project.
Relevant answer
Answer
Thank you Pawel!
  • asked a question related to Cancer Genomics
Question
4 answers
I would like to look for functional polymorphisms upstream of the transcription start site of my gene of interest. Has anyone got any experience in doing this in populations of cancer patients? 
Relevant answer
Answer
If you already have the sequences you could always align and look for SNPs, indels etc? The sample size would have to be quite large to be statistically significant, and you'd want to look at a control group (i.e. non-cancer patients) as well. There's probably quite a lot of existing sequence data out there already for a variety of cell types.
To test what functional effect these may have you could do as Jana suggests and do some Luciferase assays (or EMSA I guess), there are probably some relevant cell lines available already.
  • asked a question related to Cancer Genomics
Question
1 answer
The clonal dynamics of primary tumor cells.
Relevant answer
Answer
May be the attached articles could be of help.
Best regards
Robert
  • asked a question related to Cancer Genomics
Question
6 answers
What we know: It's been established that keap1 is involved in the regulation of transcription factor Nrf2, facilitating its ubiquitination and degradation in basal conditions (either by sequestering in the cytoplasm or actively shuttling it out the nuclei). When oxidative stress triggers a change in this mechanism Nrf2 migrates to the nuclei, where is binds the antioxidant response element (ARE), an upstreaming promoter region of many antioxidative, drug-metabolising and cytoprotective genes. This seems to be fairly well established.
But it gets more complicated than that: It appears that to do this Nrf2 requires dimerisation with other proteins containing bzip complexes. I had read that small Maf proteins were normally involved in this activation of ARE. However, I have recently come across a different model that suggests that Maf proteins actually inhibit expression of this genes and it is binding of Nrf2 to other bzip-containing factors that leads to upregulation (I've attached the model, from the NCI's cancer genome anatomy project, below).
Confusing models: Most of the literature I've found supports the opposing idea of the Nrf2-Maf dimers having an upregulating effect. Are these just different models? Has one premise been proven valid over the other? Or have I simply misunderstood this? If anyone has a better grasp of this than myself any comments would be great.
Thanks!
Relevant answer
Answer
Dr Le sirvió?
  • asked a question related to Cancer Genomics
Question
9 answers
Microsatellite instability caused by defective mismatch repair machinery is observed in up to one fifth of colorectal adenocarcinomas. When the aberrant number of short tandem repeats in the tumor cells when compared to the germline genome are detected in 30% or higher in the examined microsatellite loci, it is referred to as microsatellite instability-high (MSI-H). I find no good explanation so far to explain why MSI-H tumor is more likely originated from the right-sided colon whereas, left-sided colon got more microsatellite stable (MSS) tumors? Strange.
Relevant answer
Answer
Hi Evripidis, the developmental epigenetic signature of a tissue is very strong and particularly involves changes in the methylation of transcription factors and homeobox genes.
  • asked a question related to Cancer Genomics
Question
3 answers
How can I download the data set/sequence for chromosomal breakpoints in human cancer Genome?
Relevant answer
Answer
Thanks for your nice respond Mr. Aleksandar Sokolovski. Actually i want to deal with Hi-C data and if you have some experience please let me know. Thanks. 
  • asked a question related to Cancer Genomics
Question
2 answers
I am interested in calculating differential expression of genes for tumor vs. normal samples from RNASeq V2 level 3 datasets for TCGA (downloaded from UCSC Cancer Browser). This data is from *.rsem.genes.normalized_results, log2(x+1) transformed and upper quantile normalized
 After browsing a lot of literature, blogs and forums, I am able to note following methods of calculate differential expression from TCGA datasets.
1.     Fold change Since the values from UCSC are already log 2 transformed, for a gene x, of tumor y, Log 2 fold change can be calculated as (log 2 transformed gene expression value of tumor sample) -  (log 2 transformed gene expression value of matched normal sample).
 2.     Packages like limma to fit a linear model to the log2-transformed data using an empirical Bayes method to moderate standard errors.
 3.     Package EBSeq, as RSEM’s original paper since that algorithm takes into account the stochastic nature of the RSEM output.
 4.     Or, I should get expected_counts or scaled-estimates from TCGA data-portal and then use voom() to normalize, as well as packages like DESeq, EBSeq, limma for differential expression
 5.     Z-score (if comparing tumor vs. normal) = [(value gene X in tumor Y) - (mean gene X in normal)] / (standard deviation of gene X in normal)
 I also see variation of Z-score application as
(A)   in the TCGA cases where tumor samples are more than normal like say 500 tumor samples out of which only 65 have matched normal sample
(B)   In TCGA cases where you have only tumor samples or you want to consider only tumor samples and you can compare it with data of normal tissues as from GTEx
(C)  In cases where you have only tumor samples and twisting the calculations as
Z-score= [(gene x log value – mean of log values gene x for all tumor tissues)/ standard deviation of gene x from mean of all tumor tissues]
I would appreciate you, comment or suggest or share your experience on dealing with such kind of datasets, on the following lines:
Query 1: Have I understood these methods correct? How else we can calculate differential from TCGA datasets?
Query 2. Since, Z-score assumes normal distribution and for RNASeq expression values one would not expect normal distribution, one should go for quantile normalization, but as TCGA datasets in UCSC is already quantile normalized, should these values be used directly and simply calculate fold change with that.
Query 3. Are all the variations of Z-score mentioned in point 5 valid?
Query 4: Which of the above or any other method is the most recommended method to calculate differential expression for tumor vs. normal samples from TCGA datasets and why?
Relevant answer
Answer
You may find your answers here; it is very detailed:
  • asked a question related to Cancer Genomics
Question
3 answers
Hi,
I am searching for Structural Variation caller compatible with Mate-pair sequencing data. Any suggestion will be appreciated.
Thanks
  • asked a question related to Cancer Genomics
Question
5 answers
I have tried to search the patient`s description of the colon tumors analyzed in TCGA (the cancer genome atlas), but I have not found it; could somebody give me some information about how I might obtain the tumor characteristics associated with the analyzed samples (i.e. Ras mutation status, primary tumor, metastatic tumor... )?
Relevant answer
Answer
Hi Laura,
I think you need clinical and molecular subtypes informations related to the samples. You can take a look in TCGA's site for recent publications (markers). 
In particular you can try two functions of our new package such as: 
datClin <- TCGAquery_clinic(tumor = "COAD",clinical_data_type = "clinical_patient")
datSubt <- TCGAquery_subtype(tumor = "COAD")
For further informations or collaborations you can write me (acolapri@ulb.ac.be).
Best,
Antonio Colaprico
  • asked a question related to Cancer Genomics
Question
3 answers
I have methylation data from Infinium 450K and I found out differentially methylated regions. Is there a way to find the equally methylated regions as well, particularly in R.
Recommendation for any Package with commands would be highly appreciated.
Relevant answer
Answer
Hi Tushar, 
If I kept well your question I agree with Ferdinando and Wei you can consider the !DMR, the not significant regions.
I suggest you to try the function TCGAanalyze_DMR and TCGAVisualize_volcano from our new package TCGAbiolinks.
In particular you can take a look in our NAR paper or related Bionconductor's vignette.
For further informations or collaborations you can write me (acolapri@ulb.ac.be).
  • asked a question related to Cancer Genomics
Question
5 answers
Hello every one 
I am doing RNA seq analysis. I am new in this field, I clustered my genes but I am wondering if there is away I can fellow to compare my genes to the whole cancer genome. 
I really appreciate your help
Shaima Jabbar 
Relevant answer
Answer
Could you provide a bit more information? It sounds like you're working with Human samples and that you've done a de novo assembly, is that right? If so, have you also mapped your reads to the wild-type and cancer genomes?
What kind of comparison would you like to do? Are you interested in the differences in SNPs, or perhaps expression levels, etc?
Any information you can provide will help us to give you appropriate suggestions.
  • asked a question related to Cancer Genomics
Question
2 answers
I'm trying to determine if I can do an unpaired copy number analysis in the Partek Genomics Suite (PGS) v6.6 using .CEL files obtained from CytoScanHD chips (Affymetrix). I am able to download all necessary annotation files, however I cannot find a .cnmodels HapMap file to download, which I need to generate copy number from unpaired samples. PGS v6.6 seems to require a .cnmodels file to do this and does not accept other HapMap reference files like .REF_MODEL. I have previously used .cnmodels for SNP6 and 500K arrays in PGS however I cannot find one for CytoScanHD.
I have 2 questions which I am hoping someone can help me with:
1. Does anyone have a link to a .cnmodels HapMap file for CycoScanHD which I could download?
2. Is it possible to use a different HapMap reference file other than .cnmodels to do unpaired copy number analysis in PGS and if so which reference file and how?
Relevant answer
Answer
Affymetrix also claims that their Cytoscan HD copy number values do not need paired samples, because they use an internal reference to determine copy number.  This may explain why there is no .cnmodels file.
  • asked a question related to Cancer Genomics
Question
3 answers
I want to compare levels of an isoform between prostate cancer database and a normal database. As I am not a bioinformatician, I am looking for any help on how to do so... Are there any tools or tutorials for this that can help on this issue? I am looking at genomic databases and the protein isoform is a splice variant.
Relevant answer
Answer
@Edward Moh - I am looking at genomic databases not protein databases. so I want to quantify the amount of spliced isoform based on their unique exon-exon junctions.
  • asked a question related to Cancer Genomics
Question
4 answers
As far as I am concerned, the standard coverage 50x for the analysis of highly heterogeneous tumor samples such as melanoma is insufficient and scientific literature recommends base coverage around 500x or even 1000x. However, with that coverage the cost of analysis increase enormously.
Relevant answer
Answer
There is no one coverage that is right or wrong, better or worse. The answer will almost solely depend on the question you are trying to ask of the data and what confidence level you require for detecting SNVs. You would likely require lower coverage if you were only interested in detecting more commonly mutated genes in melanoma, such as BRAF, vs. the much higher coverage required for discovering novel SNVs if for example one were interested in doing a study on tumour evolution or spatial/temporal heterogeneity.
  • asked a question related to Cancer Genomics
Question
8 answers
If not, should there be? When we look upstream from genes, we are confronted with our personally unique 'elementome'. Cancer as a mitochondrial metabolic disease (2015 Seyfried ) considers environment but focuses on one organelle.  While it is hard to argue the great science behind SMT, TOFL (Sonneschein& Soto), Bissel's stromal related studies, and Seyfried above...but they all seem downstream in that they fail to mention, downplay  the environment, or key on specific micro env.  The way I see it, it seems highly unlikely a single organelle/compartment in a cell is responsible for the root cause. Thoughts?
Relevant answer
Answer
It is established that the dysfunction of Na+/K+-pump is a common consequence of any pathology, including cancer. At the same time, it is also known that among a number of mechanisms involved in cell volume regulation, Na+/K+-pump has a central role in it. This role of pump is due to its two important properties: a) Na+/K+-pump which generates Na+ gradient on the membrane serves as an energy source for a number of secondary ionic transporters, such as, Na+/Ca2+, Na+/H+, Na+/sugars, amino acids & other osmolytes and b) Na+/K+-pump, having the highest metabolic energy (ATP) utilizing mechanism, has a great intracellular signaling role in regulation of sorption properties of intracellular structure as well as in generation of water molecules during oxidative glucose processes in cells. Na+/K+-ATPase (working molecules of Na+/K+-pump) has 4 catalytic subunits having different functional activities and sensitivities to cardiac glycoside: α1 (low), α2 (middle), α3 and α4 (high), the latter is identified only in testis. From these isoforms α1 (fully) and α2 (partly) have ion transporting functions, while α3 isoform only performs intracellular signaling function. By previous our study it was shown that dysfunction of α3 isoform-dependent signaling function controlling cell hydration stimulates carcinogenesis: So its information for thinking.
  • asked a question related to Cancer Genomics
Question
6 answers
In TCGA cancer data, some miRNA expression(RPM) values are '0'. So do I need to consider those miRNA expression or not? Can any one give details regarding miRNA expression(RPM) data.
Thanks in advance.
Relevant answer
Answer
True 0 are always something you should take with caution... I would either dig into the protocol to see if there is something causing it or just ignoring it. The cause of a true 0 might be caused by a biological event (as mentioned in the previous answer) or by the treatment of the data. Since miRNA are small RNAs, current mapper programs have trouble to handle it. You can check if all the samples have 0 for this exact miRNA. If yes, I would remove it from your dataset.
  • asked a question related to Cancer Genomics
Question
13 answers
we consider non-genetic cell phenotype plasticity as a central process in therapy resistance. We take into account theability of cells to produce discretely distinct phenotypes, switch between them without genomic alterations and inherit the new
phenotype non-genetically across cell generations (Brock et al,2009).
Relevant answer
Answer
Yes- in cell culture systems  especially  with mixed cell populations ( for example breast cancer cells, fibroblasts and myoepithelial cells) the morphology of the cancer cells will alter depending on the ratio of the cells and also the in 2D versus 3D ( i.e gels) and also if in 3D  the phenotype will be affected by the type of matrix used - coll IV v matrigel for example
  • asked a question related to Cancer Genomics
Question
7 answers
We have knockout, wide type and heterozygotic mice, only finding heterozygotic mice got tumor(4/10). I am really confused about this outcome. I wish someone give me some ideas about this heterozygote specific phenomenon.
Relevant answer
Answer
if the gene codes for dimeric molecule then it is possible that the mutation does not change the structure therefore the activity of the molecule very much compared with the normal situation but a combination of normal and mutated chains dimerise to a molecule that lacks the correct shape to work/interact properly
  • asked a question related to Cancer Genomics
Question
15 answers
Science has provided us with countless discoveries thought to potentially improve cancer outcomes. However, only a handful of them have been translated into clinical care, and at a quite prohibitive price tag (eg new generation TKIs, monoclonal antibodies, genomic testing etc). Some other, more cost effective, are yet to be fully adopted by health care providers. Among the later I would count maximizing use of metformin in patients diagnosed with type 2 diabetes and cancer. While use of metformin as an anti-tumor agent is currently tested in a concerning high (cost-wise) number of clinical trials, maximizing its benefit among patients with type 2 diabetes is yet to be a focus despite the drug being the first line therapy.
What are your thoughts?
Relevant answer
Answer
Time after time, the marketing team from pharmaceutical companies bring us some pearls like 'target therapy', to differentiate their drugs from chemotherapy (which in fact they are a class of), or ´personalized medicine´, to emphasize some predictive factor favoring a new treatment in detriment of older ones (the search for the holy graal is permanent).
Indeed, science permited us to achieve a better understanding of cancer diseases and some new therapies emerged, at a high cost due to the vast failure in translate biological knowledge into useful drugs. Meanwhile, if a drug is not 'patentable', as metformin, the system will not provide funds enough to fully appreciate its therapeutical role in a timely manner, no matter how sound would be its background (unless a slightly modified version of the ancient drug could be produced).
About your question, marketing apart, medicine always was and must be about care a person, its fears, beliefs, habits, organism and his/her disease. At every consultation and prescription, an oncologist should 'personalize' the approach to diagnosis and treatment for that single patient - and not to abstract the patient singularity to apply blindly the newest true 'discovered' in the scientific noise from medical literature.
  • asked a question related to Cancer Genomics
Question
3 answers
I am looking for a database which has some annoatations for intergenic gene polymorphims that can help me out for cancer genomics report analysis. Can any help me out
Relevant answer
Answer
You can focus on this website (http://www.scandb.org/newinterface/about.html ).Good luck!
  • asked a question related to Cancer Genomics
Question
3 answers
and is this related to number of mutations required for cancer  .
Relevant answer
Answer
All malignant cancers need to maintain their telomeres in order to possess one of the hallmarks of cancer, limitless replicative potential. In rare cases, some cancers which do not maintain their telomeres, such as a subset of neuroblastomas, spontaneously regress. Tumours have 2 methods of maintaining their telomeres; one is expression of telomerase (hTERT) and the second is alternative lengthening of telomeres (ALT), used by a minority of cancers (but up to ~20% in certain cancer types). 
In my experience,  tumours which use telomerase to maintain their telomeres have shorter telomeres than their matched normal tissue.  Whereas the remaining 10-20% of tumours, those that maintain them by ALT have telomeres 1.5-4.5x longer than those from matched normal tissue. I have seen this both in my data on various brain tumours and recently at a conference where several scientists presented data on telomere length in normal colon, polyps and colon cancer.
Telomere dysfunction in general has been implicated as a potential initiator of genomic instability, but I have not seen a study comparing if universally tumours with shorter telomeres have higher mutation rates that tumours with longer telomeres such as those that use ALT.
Various cancer types and tissues have intrinsic mutation rates and there are many cancers that have very low mutation rates and no recurrent mutations. Therefore I would be careful when stating anything about the number of mutations required for cancer as there are many cancers where mutations do not seem to be the cause. Even high mutation rates seen in certain cancers could be a late event of tumourigenesis, and perhaps initiation of telomere maintenance is a later even during tumour formation.
In the last few years there have been studies looking at hTERT promoter mutation as a potential driver of hTERT expression in cancer.
  • asked a question related to Cancer Genomics
Question
3 answers
Data from literature are conflicting concerning the question whether impaired DNA repair might explain genomic instability in CSC
Relevant answer
Answer
In general, CSCs are considered to harbor high potential to repair DNA damage under inflammation and ROS in the tumor micro-environment.  On the other hand, given that some kinds of tumors arise from CSCs with accumulation of genetic mutations, the genomic instability in CSCs is likely as compared with non-CSCs. In my opinion, the potential of DNA repair in CSCs might be different each other at the carcinogenesis and during the tumor development including invasion/ metastasis and relapse. I mean that pre-malignant cells with higher genomic instability would be CSCs, and furthermore, CSCs later acquire rapid DNA damage repair potential.  
  • asked a question related to Cancer Genomics
Question
3 answers
I am looking for a database which has some annoatations for intergenic gene polymorphims that can help me out for cancer genomics report analysis. Can any help me out?
Relevant answer
Answer
The UCSC genome browser has some SNP annotations, take a look if it has what you need http://genome.ucsc.edu/
  • asked a question related to Cancer Genomics
Question
2 answers
Can anybody suggest me a technique to measure genomic stability as histone modification read out. I am trying to measure the stability of cancer genome, what techniques can I use to measure the genomic instability in live cells, FFPE samples. Trying to establish an epigenetic signature for stability, what techniques could be of use here?
Relevant answer
Answer
Are you looking at genomic instability in the presence or absence of certain histone marks? One thing that comes to mind is looking at microsatellite instability. If you're willing to share a more detailed hypothesis, myself or others might be able to suggest more appropriate techniques. Cheers.
  • asked a question related to Cancer Genomics
Question
9 answers
I work with a protein that has 2 different common SNP variants in the population, and we discovered that only one form can bind a protein required for degradation. Is there a resource that links cancer panel genome wide SNP (DNA sequence) with genome wide protein expression level (eg mass spec, protein or IHC microarray) from the same samples? Or do I have to find a patient/control set and do it myself?!
Relevant answer
Answer
Have you tried TCGA??
  • asked a question related to Cancer Genomics
Question
4 answers
i am checking for TSG (p16,p21,PTEN,p53) Methylation, mutation status of ovarian cancer cell line like oaw42,oaw28 ,ov167,ov177
Relevant answer
Answer
Thank you! 
  • asked a question related to Cancer Genomics
Question
4 answers
I have some old CGH data from Roche/NimbleGen that I would like to organize into some publication quality figures. I find the software supplied by them to be inadequate for anything but scanning the results.
Relevant answer
Answer
I like Nexus but it's expensive.  A free option is to convert your data into .seg format and load it into IGV: http://www.broadinstitute.org/igv/SEG
You can then export the plots in .svg format to make nice figures.
  • asked a question related to Cancer Genomics
Question
1 answer
The problem in downloading from FTP is that it provides Genbank file of whole genome. How to extract coding region from these Genbank file?
Relevant answer
Answer
I would suggest the table browser of UCSC (see link).
I added an preview of how to get the sequences for the RefSeq gene sequences form hg19.
This should provide you with the fasta sequence of just the coding regions of the genome of your choice.
>hg19_refGene_NM_032291 range=chr1:66999825-67210768 5'pad=0 3'pad=0 strand=+ repeatMasking=none
TTTCTCTCAGCATCTTCTTGGTAGCCTGCCTGTAGGTGAAGAAGCACCAG
CAGCATCCATGGCCTGTCTTTTGGCTTAACACTTATCTCCTTTGGCTTTG
etc.
  • asked a question related to Cancer Genomics
Question
3 answers
Is it a target or over expressed during carcinogenesis?
Relevant answer
Answer
Thank you so much providing such an important article.
  • asked a question related to Cancer Genomics
Question
10 answers
Does anyone know a way to identify the main GO Terms associated with cancer?
I have a list of genes annotated with GO biological processes and wanted to filter only those annotated in biological processes related to cancer.
Relevant answer
Answer
I don't think you can find GO terms which are associated with cancer because you could pretty much associate almost any GO term with having some relation to cancer. There are no specific cancer GO terms. All are related to some normal cell function that you could argue are dysregulated in cancer like angiogenesis, cell cycle checkpoint or kinase activity to name a few.
I think your best bet is to search for cancer associated genes here (COSMIC database) : http://cancer.sanger.ac.uk/cancergenome/projects/census/
...and then filter your gene list based on this database.
  • asked a question related to Cancer Genomics
Question
12 answers
I want to use the cancer RNA-seq data from TCGA to do some further study but I have no idea to download those NGS data.
Relevant answer
Answer
Hello,
In order to download data from TCGA data portal:
2. Select the cancer subtype you are interested in (i.e breast invasive carcinoma)
3. Select mRNA
4. Now you can see a table where rows are representing different patients.
5. If present select the column (by clicking on header) that referse to RNASeq or RNASeqV2 if it is present for that cancer subtype and then click BUILD archive.
6. Keep in mind that just below the header there is a number indicating the respective data level. Levels 1-4 (https://wiki.nci.nih.gov/display/TCGA/Data+level)
If you need RAW data such as FASTQ files you have find level 1 data, but often this kind of data is not publicly available on TCGA and you might need to ask for permission in order to download it.
Hope I helped you somehow,
Cheers,
Marco
  • asked a question related to Cancer Genomics
Question
4 answers
This project is in a rare type of lymphoid cancer. Due to this rarity, there are only 3 cell lines available for the disease, which I have in the lab. They are drug-sensitive cell lines (drug disclosed only if you are interested in collaboration) and I also have their 3 drug-resistant clones, which were developed in the lab (so total sample n=6). Unfortunately, I do not have any patient samples. I wish to conduct RNA-seq to examine changes at the mRNA level as well as SNV's in these cells. The goal is to identify changes that characterize drug-resistant cells vs. their drug-sensitive WT counterparts. Please tell me if this type of an analysis is possible with such low sample numbers? If so, please let me know if you are interested in collaborating to analyze the resultant RNA-Seq data?
Relevant answer
Answer
Dear Aneel,
First of all, I would like to tell you that it is possible to perform this kind of analysis with such a low no. of samples. Actually you want to measure differential gene expression (DGE) which is a quantitative experimental set-up for RNA-seq. Here biological replicates are important and 3 vs 3 will be enough to perform this kind of statistics which we call Explorative or as discovery setting. However, latter on you need to validate your findings on larger cohort which you already know.
I have to admit the we don't work on Lymphoid cancer, so I can't offer any possibilities of collaboration. But my point is to clear your doubt about sample size for RNA-seq...I would say go for it...
Hope this would help you...good luck for successful collaboration
  • asked a question related to Cancer Genomics
Question
25 answers
I would like to perform a in-silico validation for my research study where I need to combine some published datasets (from GEO portal) to increase the number of samples (n) and then analyse them for differential expression for specific genes. I would to do all these analysis in R. Any suggestion for R software package, combining package or batch effect removal + their r scripts?
Thanks a lot for anticipation
Relevant answer
Answer
As @Rekins suggested InSilico DB has a "merging" R-Bioconductor package to combine public datasets from GEO. If you are not using R you can also combine data from the online platform (https://insilicodb.org)
Example:
# Retrieve 2 datasets
eset1 = getDataset(gse="GSE10072", gpl="GPL96", norm="ORIGINAL", genes=TRUE);
eset2 = getDataset(gse="GSE7670", gpl="GPL96", norm="ORIGINAL", genes=TRUE);
#combine them
esets = list(eset1, eset2);
eset = merge(esets, method="NONE");
#plot them
plotMDS(eset, targetAnnot="Disease", batchAnnot="Study");
InSilico DB packaged various batch removal effects methods so line 4 could be replaced with:
eset = merge(esets, method="XPN");
or
eset = merge(esets, method="COMBAT");
Hope this helps.
For more info Bioinformatics paper reference; InSilico DB and InSIlico Merging packages links, and blog link.
- Unlocking the potential of publicly available microarray data using inSilicoDb and inSilicoMerging R/Bioconductor packages -BMC Bioinfomatics [http://www.biomedcentral.com/1471-2105/13/335/abstract]
- inSilicoDb: an R/Bioconductor package for accessing human Affymetrix expert-curated datasets from GEO - Bioinformatics [http://bioinformatics.oxfordjournals.org/content/27/22/3204]
R-Bioconductor packages:
and
  • asked a question related to Cancer Genomics
Question
1 answer
There are controversial questions about the role of viruses in some types of cancer. There are some viruses "associated with" or "suspected to play a role in" some types of cancer. The major problem is that there is a lack of evidence for the presence of the virus in the majority of the observed cases leading to the conclusion that the virus cannot be considered as the cause of cancer.
So my question is simple, can we certify that there is an absence of a virus in a cell with today's techniques or are we limited with a detection threshold if a small number of copies may be sufficient to trigger cancerogenicity?
Relevant answer
Answer
There are a few categories of answer to your questions:
[Theoretical Answer] You only need 1 copy of an RNA to detect it, providing the sequencing depth is sufficiently high, the enrichment for RNA doesn't preclude the viral RNA (like oligo-dT priming might), and you are aligning your RNA-seq to a viral genome. If you set up spiked in controls, you could detect it providing you know what you are looking for.
[Practical Answer] A virus that is causing cancer will usually be pretty abundant. A paper by my colleagues Aparna Bhaduri and Kun Qu developed software to detect viruses in RNA-seq (see here for details about how many reads were needed, etc.: http://bioinformatics.oxfordjournals.org/content/early/2012/02/28/bioinformatics.bts100). For practical purposes, I'd say that if you know what you're looking for, you can make a yes/no call with reasonable accuracy on the presence of a virus.
[Other] Even if you detect a million copies of an RNA virus in the cell, it doesn't matter if the virus has an unknown sequence and you aren't aligning to it, or even if you do de-novo sequence assembly, and don't know what the virus does. There are likely many viruses that have an unknown effect and you need to do a Koch Postulate-Style (http://en.wikipedia.org/wiki/Koch's_postulates) set of experiments to see if that was indeed your oncogenic event.
  • asked a question related to Cancer Genomics
Question
3 answers
When does the term transcriptome come in case of the cancer cells cultured in vitro? If we are thinking for similarity determination of the cancer cells in comparison to original tumor tissue from which it has been derived the early passage, cells are considered for studying NGS data because there is some common hypothesis and prove for some cancer cell lines that after some passages (5-15) genomic and transcriptomic alteration occurs. I am not sure in between the early passages, which passage will be utmost suitable because NGS run is always associated with high costs.
Relevant answer
Answer
tl;dr: qRT-PCR first to see if contaminating cells are gone
If I am reading your question correctly, I'd suggest that you profile a cell line derived from a primary tumor at the earliest passage possible where you no longer have contaminating (stromal cells, immune cells, etc.) cells also in the culture.
Even a small amount of contaminating cell infiltrate will have highly abundant gene products that will complicate your classification against different tissues when comparing after your NGS run (I assume we are talking RNA-seq here).
Experimental suggestion: if you are taking a skin cancer biopsy that should only contain epidermal keratinocytes (the cell of origin for squamous cell carcinoma, let's say), design some qRT-PCR primers for genes exclusively found in an immune system cell, a pigmented melanocyte cell, and others in the skin. When the detection of these genes has fallen significantly from your original culture, I'd say you are ready to sequence the transcriptome for a relatively pure, early passage cell line.
  • asked a question related to Cancer Genomics
Question
9 answers
Genome sequencing of tumors could help in dissecting clones within a tumor, is that possible to predict drug targets for each clone, so that a cocktail could be applied to the patient?
Relevant answer
Answer
Solid tumors are not always monoclonal. A number of studies indicate intestinal tumors from mice and humans can be polyclonal origin. Thus, some tumors can be heterogeneous very early during tumorigenesis. However, the elimination of one clone could potentially impact other clones if interactions among different populations are necessary as tumors form form and grow.
  • asked a question related to Cancer Genomics
Question
5 answers
My project involves screening for novel SNPs in tumor samples. I've identified a novel SNP locus in a particular gene. How can I proceed to design primers for PCR testing on a bigger tumor sample? Is there any easy to use software to do that?
Relevant answer
Answer
Hi Jason,
If you know the coordinate of your SNP, i.e. chromosomal location, you can go Primer-Blast Tool in NCBI website and add a region upstram and downstream to your target and then choose the parameters (product size)
So it is better to add nucleotides upstream and downstream for your gene from (Change region Shown) and click (update view), so the target region now is getting bigger depending on how many nucleotides you added to it.
From (Primer Pair Specificity Checking Parameters Specificity check) You need to be sure to select the right choice depending on your application and if you are working on humans. If so please choose (Genome, reference assembly for selected organism) and the number 9606 (human) will appear for the organism.
Finally, click on GET PRIMERS.
This is a direct way so you don't need to worry about having a software.
  • asked a question related to Cancer Genomics
Question
7 answers
The most used definition for "Neoplasia" was put forth by Willis RA in 1952 based upon the understanding of the disease process of that era. Since then our understanding of neoplastic process has increased so much.
1. Do you think even after 60 years, the definition is still valid or needs modification? For eg., the hall mark features of cancer proposed by Douglas Hanahan and Robert Weinberg in 2000 conveys much better fundamentals than old definition.
2. If so, has any agency such as IARC has taken steps to define an universally acceptable definition that encompass all current knowledge of the process ?
3. Do you think this exercise is a waste of time and we need to focus efforts for more in depth understanding of the disease than on defining the disease?
Relevant answer
Answer
Yes Friends as the neoplasia is a an extremely complex disease, to find a suitable definition seems to be a herculean task. However, we should look for a broad and short definition , which may encompass most, if not all features of neoplasia.. In this context, 'all those who matter in oncology' should sit together and find a suitable answer.
  • asked a question related to Cancer Genomics
Question
2 answers
When everyone is thinking of bad about cancer, I wish to know much about my friend "Cancer" who is the nexus of the Healthcare business and emerging burden of the society. I would like to know some goodness of the cancer cells. I hope this will add a caliber for some of the research hypothesis with the help of the scientific society.
" When one door closes another opens"
Yours suggestions and comments are highly appreciated
Relevant answer
As far as Biology is concerned, cancer cells are amazing model systems to study the role of mutations, gene regulatory networks and signaling pathway mechanisms. Our understanding of cellular and molecular biology. The complexity of a cell's regulation and function can probably be better understood through a cancer cell.
  • asked a question related to Cancer Genomics
Question
3 answers
In NGS, scientist go through different processing steps using standard tools and public databases to align sequences, map then on reference databases and identify different mutations known to be relevant in particular diseases. I would like to survey about the most widely used tools and reference databases.
Relevant answer
Answer
I'm guessing most folks follow the "best practices for variant detection" guidelines described by the folks who wrote the GenomeAnalysisToolkit (found here: http://www.broadinstitute.org/gatk/guide/topic?name=best-practices) . BWA is probably the most commonly used alignment tool, but some folks may use MAQ, SOAP, or Bowtie. Additionally, some people will use samtools / mpileup instead of the GaTK to actually call the variants.
HTH
  • asked a question related to Cancer Genomics
Question
1 answer
Does anybody have experience with (or know of anybody) performing whole genome sequencing (WGS) of genetically matched (i.e., from the same individual) peripheral blood leukocyte vs lymphoblastoid cell line samples as constitutive tissue controls for tumors WGS? Unfortunately, we are in a position where we do not have matching PBL for all tumor samples, but (fortunate to?) have immortalized LBCLs. How good of a surrogate are the LBCLs for such studies if no PBL Is available? In addition to SNVs I am particularly concerned about possible differences in CNVs.
Relevant answer
Answer
For analyzing the data, a lymphoblastoid cell line would not be a useful control for WGS; you wouldn't be able to subtract out all the unimportant SNVs in the patient's germline DNA. If you can get any other tissue from the patient to use as germline, that would work -- it doesn't need to be the same tissue if you're not doing a gene expression study.
  • asked a question related to Cancer Genomics
Question
9 answers
Can anyone tell me how to create a heatmap using circos software? I also wish to know the input format to be fed in the software. Anyone who has the knowledge on Perl can really help me in teaching this software.
Relevant answer
Answer
it will not paint a heatmap of array data, instead it paints quality measures of affymetrix data, however, it might get you a starting point for your task.
hth