Skip to content

About

I (Ethan Ford) got my start in molecular biology at the University of California, Santa Cruz when I joined the lab of Dr. Manuel Ares, Jr.  At last, after all those years of reading text books, I was finally doing real science creating a RNA cyclase ribozyme from an inside-out group I intron.  I left Santa Cruz for Cold Spring Harbor where I did my Ph.D. work under the supervision of Dr. Nouria Hernandez studying the mechanisms of snRNA transcription.  In those days before genomics, long cold days in the cold room purifying proteins seemed to be where it was at.  Next, I spent a short year at Oregon Health Sciences University in the group of Dr. Buddy Ullman where I worked on mechanisms of drug resistance in Leishmania.  Then it was off to MIT with a NRSA fellowship where I joined the group of Dr. Leonard Guarente to focus on the emerging field of the genetics of aging.  Finally, I left for Greece with a Marie Curie fellowship to study how NAD+ metabolism effects chromatin structure and gene expression in response to virus infection.  In this journey I have gone from molecular biologist to protein biochemist to microbiologist to yeast geneticist to cell biologist to computational biologist.  I am currently focusing on an ‘omics’-based project using a combination of next-generation sequencing (genomics) and immunoprecipitation-mass spectrometry (proteomics) with a some cell biology and cytogenetics thrown in.

I can be reached by e-mail at: eford(dot)dna(at)gmail(dot)com

13 Comments
  1. Alex Clop permalink

    Dear Dr Ford,

    I did some ChIP on mono-nucleosomal chromatin for different histone modifications in T cells and would like to pool some samples and run them in a HiSeq2000 system. For that I need to index the samples and I found your protocol using the TruSeq DNA library Prep kit. In your protocol you use NEB reagents and not any part of the ChIPseq kit.

    My plan is to use the ChIPseq prep kit and protocol but replacing the adapters and oligos (i.e., “discard” the ones from the ChIPseq kit and use the ones from the TruSeq DNA kit). One of the problems I am facing is the concentrations, sequences and sizes of the TruSeq oligos as these are proprietary and Illumina do not disclose them. However, I saw you talk about 0.25 microM and 25 microM for adapters and primers, respectively. Did you get that from Illumina’s Tech Support or you quantified it yourself (Bioanalyzer amount and size measurements and molarity calculation). If I have to calculate it myself, I am not sure whether the modifications on the oligos would affect these measurements.

    Another think that worries me is the annealing temperature during the PCR steps as TruSeq recommends 60C, ChIPseq recommends 65C and you use 63C, and of course, this depends on the primers used but also on the Taq and buffer used, which I cannot compare. I was wondering whether you could comment on it.

    Also for the PCR, you split it in 2 PCR to convert primers to dsDNA and avoid weird migrations. How necessary is this? I would rather go ahead with the 18 cycles recommended in the ChIPseq protocol but I am not sure whether this would mess my experiment.

    I really appreciate your comments and your time.

    I am looking forward to hearing from you.

    Sincerely,

    Alex

    Alex Clop
    Dep of Medical and Molecular Genetics
    Kings College London
    9th floor Guy’s Tower
    Great Maze Pond
    SE1 9RT London
    United Kingdom

    Tel.: +44 (0)2071889505
    e-mail: alex.clop@genetics.kcl.ac.uk

  2. ethanomics permalink

    Hi Alex,
    You have a few questions so I’ll try to address them one by one.
    1) I’m pretty sure the reagents in the Illumina ChIP-seq kit were made by NEB. This has probably changed since they acquired Epicentre, but regardless they will work.

    2) The adapter concentration is more or less 5-fold (?) more then the starting material concentration which seemed like a reasonable amount to use. As with any library preparation protocol, you should adjust the amount of adapter you use to the amount of starting material and not rely on ‘stock’ concentrations provided by Illumina. I also tried a couple concentration and went with what worked best. Look through some of my posts and there is the concentration on the TruSeq adapters in the current kits.

    3) The concentration of the PCR primers is what Illumina was using in their PCR primer cocktail. Regardless, the concentration here is not very critical.

    4) The annealing temp for the PCR seems to be pretty flexible. 63˚C works well, so does 60˚C. I have not tried 65˚C. The PCR primers in the Illumina ChIP Seq kit are very long so you can go with a much higher melting temp, probably even 72˚C. My PCR primers are different then what Illumina uses. Illumina most likely uses the TruSeq Universal primer as one primer and the other is most likely something like what I am using. But I’m guessing here since it is proprietary information.

    5) I recommend that you do not use Taq DNA polymerase. Phusion is good but the Kapa HF polymerase is better.

    6) The first short PCR is absolutely essential. The Y-shaped adapters migrate very strangely in an agarose gel so you must convert the DNA to double stranded DNA before gel purification. It wasn’t such a big deal with the old adapters since they were much shorter.

    7) I highly recommend you follow the quantification procedure I outline to determine the number of cycles to use in your PCR. Blindly going to the ‘recommended’ 18 cycles will likely result in over amplification of your library. Not such a big deal but not ideal.

    Finally, the protocol really does work well. Make sure you use the Ampure XP purification at the end.

    Good luck,
    Ethan

  3. kalyankpy permalink

    Hey, I have found a person whose passion and interests go around the NGS (wet and dry). I am on my way to gain some experience in this field. Hope we share our discoveries further with the same enthu! Nowadays I am working on RRBS! I have done the preliminary data screening! Will share about my experience in the coming days!

  4. ethanomics permalink

    Hi Kalyankpy,
    I’m starting some RRBS-seq and MeDIP-seq experiments. I’ll post more on this some time in the not so distant future as I become more confident in my expertise. I think I have a good idea to improve MeDIP-seq methodology, but the protocol is still a ‘Beta’ version. When my Covaris microTubes come next week, I’ll give it a go. What are you using to align your RRBS-seq data?

    • Kalyan permalink

      Hi ethan,

      I have assumed that whenever you reply in the blog, I will get an email. So, I kept waiting for an email. Somehow I came back to you blog to see that you have replied long ago!

      I have done the generalized (whole genome and not gene specific) analysis of the RRBS data for four samples. I have already shared my observations as comments to your recent post!

      Here is what I have done:

      1. Extract the sequences using CASAVA 1.8+ (Sanger quality)
      2. Do a quality trimming (Condetri)
      3. Adapter trimming (Cutadapt – I find this to be very flexible)
      4. Map with Bismark (This runs bowtie in 4 instances at one go. Thus requires good amount of RAM)
      5. Extract the methylation information from Bismark output (bismark scripts)
      6. Convert the methylation info into bedformat (I tweaked the scripts provided with bismark to generate the bedformat)
      7. Visualise in UCSC.

      I can personally share the tracks if you are interested in visualising!

      Have a good time with RRBS!

  5. Hi Ethan and everyone,

    I did prepare five indexed libraries with my ChIP samples from mononucleosomal (about 150bp) chromatin. H3k4me1 IP (Sample1), H3k27Ac IP (Sample2), and 3 inputs (Samples 3, 4 and 5). Sample 3 is the same original Chromatin as sample 1 and 2. I used 10 ng for the Sample2 (all the amount I had), and 20 ng for all the other samples.

    I used the ChIPseq protocol for the end repair and the adenylation and the TruSeq DNA library prep v2 protocol for the ligation and the PCR. The only modifications relate to the TruSeq protocol and are:
    (1) I dilluted the adapters in H2O 1:40.
    (2) As suggested by Ethan, I did first a 4 cycle PCR, followed by the gel size selection (cut an invisible band between 200 to 300 bp according to the ladder) and Qiagen MinElute gel purification, and a final PCR of 14 cycles. Both PCR were as suggested in the TruSeq protocol ( similar to the ChIPseq but with 60C instead of 65C annealing).

    Only Sample2 worked, the only one I processed with 10ng starting DNA, showing a peak of 270 bp (150 insert size and 120 adapters) and all the others did not. On the Bionalazyer HS DNA chip I can see that two of the input chromatins have a strange ladder with several peaks differing 30 bp between each other and ranging from 230 to 570 bp, which I am unable to interpret. Because I cut a band of 200-300bp I assume that this ladder has to come from the 14 cycles last PCR but still I don’t know what could it be…primers concatenating???? weird!!! I will be happy to attach the bioanalyzer pdf file but I am not sure this tool is possible, but I can send it by email.

    I would appreciate any comment or suggestion on what could that ladder pattern be, why most of my samples would not have worked, and how to proceed from here.

    Merry Xmas to everyone

    Thanks Ethan for your fantastic web resource.

    Cheers,

    Alex

  6. ethanomics permalink

    Hi Alex,
    I’m sorry to hear about your library prep. Together with the ChIPs, that is a lot of work. Did you do something to anger Santa Clause? My first advice would be to follow my protocol because it works. When you start putting together different things you usually have to do some optimization so it would have been better to try it out with some input sample first. That being said, I think there is a chance your libraries are fine but ‘over amplified’. I don’t really know what over amplified libraries look like with the TruSeq primers since I started doing a qPCR to get the number of cycles correct before I switched over the the TruSeq primers. At this point I would suggest taking your libraries (or at least one of the input libraries) and add fresh primer (twice the amount you used to do your PCR) and 2x PCR master mix. Do exactly one cycle of PCR, i.e. heat to 98˚C x min (x depends on the polymerase you are using), 60˚C for 30 seconds (if you are using Illuminas TruSeq PCR Primer Cocktail) and 72˚C for 90 seconds. Rerun your samples on the Bioanalyzer, if you get back a single band the your multiple bands was from overamplification. See this post:
    https://ethanomics.wordpress.com/2011/11/17/larger-then-expected-fragment-sizes-during-illumina-library-preparation/

    I would also suggest you post your Bioanalyzer image on SeqAnswers. There is an impressive amount of knowledge on that forum.

    Glad you like my blog!!!!

  7. Alex Clop permalink

    Thanks Ethan. I posted it in SeqAnswers as you recommended (posted on 01-09-2012, 05:49 PM).

    Santa Claus will have to be nicer with me in 2012 🙂

    Also, as you said, I will try this additional PCR to test for and correct possible over-amplification.

    Since I already bought the ChIPseq and the TruSeq DNA v2 library prep kits I would like to do a bit of optimization in input samples. I will also do a reaction without DNA to see whether the ladder-like peaks form as well.

    To be honest with you, your protocol is a very good option but I am wondering whether it would more effective for me to switch to the Bioo NEXTflex™ ChIP-Seq Kit and barcodes as altogether cost $500 and I guess that purchasing the reagents needed for your protocol would cost more. Any comment on that? For what I have read so far, it seems that the users of the Bioo kit are satisfied too. I would appreciate any comment from the people that have used it.

    Wish a great 2012 to everyone.

    Alex

  8. ethanomics permalink

    Hi Alex,
    I came up with my protocol because the only method I could find at the time was putting the bar code at the end of the adapters so the first 4 sequencing reads were the bar code. This has a lot of disadvantages over using an index read so I had to come up with something. If the Bioo NEXTflex kit was available, I probably would have used that. The other thing is I already had all the enzymes so it was just buying a couple oligos. My point is for you the Bioo kit would definitely would be a good choice. I’m not switching to it because I like that the Kapa polymerase is very efficient and I can pipe the same reagents into other methods like MeDIP-seq and RRBS-seq.

    You should very easily be able to use the reagents from the TruSeq DNA kit for ChIP-seq. Just make sure you do either the entire PCR or at least a few rounds of PCR before you gel purify, which is what is done in my protocol and the Bioo protocol. Also, make sure you do a qPCR to determine the number of cycles to amplify your library (see my protocol for details). This is important for all protocols. It adds 50 min to the procedure but it is well worth it!!!! Also since you are looking at mono-nucleosomal DNA which is only 140 bp, I would either use Qiagen minElute columns or 2x volumes of AMPure beads. Size selection by AMPure with a 140 bp insert I can’t see working very well. I guess what I am saying is follow my protocol using the enzymes from your ChIP-seq kit and/or TruSeq kit. You don’t have to order anything.

    Taking a look at your bioanalyzer image, I would bet that that ladder is ‘over amplified’ adapter-dimers. Not something you would want to sequence and not worth spending more of your time with.

    Ethan

  9. Hi Ethan,

    I just wanted to drop you a comment to say how thankful my group is that you have provided all these insights into library prep. You are often referred to as our “good friend Ethan” who says to do this then that. Personally, I think you should patent your pre-PCR quantitation idea. We dropped 7 cycles off the NEB CHIP-seq kit recommendations, and 5 off the Illumina mRNA-seq protocol (that one can likely go lower yet, probably 8 cycles on 2ug inputs).

    If someone is smart out there they should be trying to hire you ASAP in my opinion

  10. Alex Clop permalink

    I am of the same opinion. Thanks Ethan for your fantastic motivation and for sharing with us your knowledge and experience.

  11. ethanomics permalink

    Hi John,
    Thank you very much for the kind words and encouragement. Sharing is what science is all about. We figure out stuff, tell people about it and that’s how science moves forward. If you don’t tell anyone, you have an impact factor of zero. Old school journals are not the only game out there. And, BTW, your thread on SeqAnswers “I use to think I was good at a computer” is probably the best thread on the whole forum and thanks to you for your scripts.

    Ethan

  12. Ming Dong permalink

    HI, I almost decided to use desalted homemade oligo as adapter until this paragraph from Sigma freaks me out:

    http://www.sigmaaldrich.com/life-science/custom-oligos/custom-dna/learning-center/best-purification.html

    PAGE
    The basis of the PAGE separation… resulting in purity levels of 95–99% full-length product. …PAGE is the recommended purification for longer oligos (≥50 bases).

    I am using Illumina Adaptors, so ~63 bp. I am sure you got good results with desalted (non-PAGE purified) oligos, right?

    How long are your oligos?

    thanks

    Ming

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: