Bioanalyzer indispensable?

[Sandokhan]

Hi, at my lab we are going to buy a PGM, plus almost all the recommended equipment, one-touch, qubit, server, pippin and minifuge. As we run out of money, no bioanalyzer or sonicator. The question(s) is, is the bioanalyzer really essential? Are we going to waste our money on a very expensive door stop? Can we get away by using other methods that do not require the bioanalyzer?
Thanks for any advise on this question from a novice.

[ECO]

In my experience, especially during the early stages of getting an NGS system up in new hands, the bioanalyzer is an indispensable piece of QC equipment.

You'll find it valuable to check libraries post shearing, post ligation, post size selection, post PCR, etc. You could do all these on gels, except perhaps quantitate accurately. Once you get the protocols dialed in for your applications, you may find it less necessary.

[pmiguel]

I disagree with Eric. We use our Bioanalyzer heavily. But I am far from certain it is a necessity. In fact, it is a rather cantankerous instrument--easily 10% of lanes have bizarre spikes and bulges that disappear when repeated. The reagents are expensive and in some cases go bad even before their expiration date. Their tech support is incapable or unwilling to detail basic, non-intuitive, issues with the function of the instrument. (For example this and this.) Most of what it does can be emulated with minigels and decent gel documentation software.

To be fair, like I said above, we use ours heavily. But if I were starting again I would seriously consider a Caliper instrument instead. And if I were starting a PGM core -- or any other core where $20K was a sizable percentage of the cost of my start up, I would strongly consider doing without.

--
Phillip

[HESmith]

One huge advantage of the Bioanalyzer over gel electrophoresis is sensitivity, which allows you to visualize the size distribution and degree of adapter dimer contamination of your final libraries even at dilute (< 1ng/ul) concentrations. I agree with the points that Phillip raised; the instrument is finicky and tech support is lousy.

-Harold

[Sandokhan]

Thanks for the info. I know having the bioanalyzer is desirable, but sadly our budget for equipment was cut down, so, we'll have to manage without it and learn the hard way.

[aleferna]

We are also considering the IonTorrent but what about the DNA fragmentation?

So there is no way in hell we are buying a Covaris(its too expensive!), we've been using Nextera for this. Does anybody know where I can find a comparison between Covaris, Nextera and the new IonShear kits??

[ECO]

Their tech support is incapable or unwilling to detail basic, non-intuitive, issues with the function of the instrument. (For example this and this.) Most of what it does can be emulated with minigels and decent gel documentation software.

Just saw this...

With regard to your two examples of "basic" issues...I'm not sure why you expect dsDNA and RNA to run at the same mobility on an RNA chip...and then expect Agilent to support it? Did you denature the dsDNA? Do you know the dye-binding affinity for ssDNA versus dsDNA versus RNA? There is almost certainly a different ratio of dye:nucleic acid for RNA vs DNA which could totally explain the mobility difference, let alone how the software works to normalize even the mobility times shown.

The size calculations for your sample are based on the mobility of the ladder, and normalized to the marker peaks included in each lane. It's essential that all those molecules are of the same general mobility in the gel matrix. It's an exercise in futility to try to size dsDNA against an RNA ladder, and the converse, as the ladder and the marker peaks will migrate differently from your sample.

There is a lot more going on from a biochemical standpoint than you appear to be giving Agilent credit for. Measure DNA on a DNA chip with a DNA dye against a DNA ladder with DNA lane markers. Measure RNA on an RNA chip with an RNA specific dye against an RNA ladder with RNA lane markers. OR include your own standards, make your own measurements for standards, and calculate your own molecular sizes. Neither of the observations you've made in your two threads are surprising in the slightest.

To address the original poster's question, I still feel that when you take all the parameters into account (time, data quality, reproducibility, instrument cost, etc), the bioanalyzer performs very valuable functions for an NGS lab. If you have a dialed in protocol for one sample type/size, you definitely can get away without it. If you ever expect to troubleshoot library prep, shearing, or similar, it's great.

[pmiguel]

Just saw this...

With regard to your two examples of "basic" issues...I'm not sure why you expect dsDNA and RNA to run at the same mobility on an RNA chip...and then expect Agilent to support it?

I don't expect them to run at the same mobility. I expect single-stranded stuff to run faster than double-stranded molecules of the same length. I am also okay with them not running at the relative mobilities that I expect, I just want Agilent to specify what their relative rates are -- or at least mention that they may be non-intuitive.

Actually, I now know roughly know what their relative rates are, so I just want Agilent to provide that information to all their customers. That is the level of support I expect.

Did you denature the dsDNA?

Yes, when I run it on an RNA chip, I do. Are you sure you read those posts?

Do you know the dye-binding affinity for ssDNA versus dsDNA versus RNA?

How would I know that? Agilent does not divulge what dyes are being used.

There is almost certainly a different ratio of dye:nucleic acid for RNA vs DNA which could totally explain the mobility difference, let alone how the software works to normalize even the mobility times shown.

I am fine with the BioAnalzyer not running exactly like an agarose gel. I just want Agilent to inform me about it.

By the way, dye:nucleic acid ratio differences between ss and ds are a reasonable hypothesis about the unexpected relative migration speeds. But why are we hypothesizing about something Agilent will almost certainly know for sure?

The size calculations for your sample are based on the mobility of the ladder, and normalized to the marker peaks included in each lane. It's essential that all those molecules are of the same general mobility in the gel matrix. It's an exercise in futility to try to size dsDNA against an RNA ladder, and the converse, as the ladder and the marker peaks will migrate differently from your sample.

Maybe. But that isn't really the point. The point is that you can end up with ss molecules in a ds sample (or vice versa). Unintentionally. (We are generally using these for QC, right?) If you are looking at a weird result and trying to figure out what is going on, don't you want to know that ss stuff is going to run slower than ds stuff of the same length? Seems like that would be helpful information.

There is a lot more going on from a biochemical standpoint than you appear to be giving Agilent credit for. Measure DNA on a DNA chip with a DNA dye against a DNA ladder with DNA lane markers. Measure RNA on an RNA chip with an RNA specific dye against an RNA ladder with RNA lane markers. OR include your own standards, make your own measurements for standards, and calculate your own molecular sizes. Neither of the observations you've made in your two threads are surprising in the slightest.

Writing that they don't "surprise" you is not the same as writing that you expected this to be the case. Did you? Not that it was something you could believe, but that it was something you already expected?

To address the original poster's question, I still feel that when you take all the parameters into account (time, data quality, reproducibility, instrument cost, etc), the bioanalyzer performs very valuable functions for an NGS lab. If you have a dialed in protocol for one sample type/size, you definitely can get away without it. If you ever expect to troubleshoot library prep, shearing, or similar, it's great.

Okay, I'll admit that I was sublimating some previously suppressed rage at Agilent over various minor issues I have had. Also, yeah, if you have a half million dollar sequencing system it would be deranged not to pick up a $20K bioanalyzer along with it. But we are talking Ion Torrent here. Less than 1/5th the cost of traditional next gen systems to have running in your lab.

Do you really need it? I don't think so. But I will admit that this is easy for me to say because we have one and use it frequently -- daily, on average.

Also, to be fair to Agilent they have already contacted me with regards to the post above and asked for clarification about complaints I had. So it isn't like they don't put any effort into dealing with customer complaints. I replied with my laundry list of stuff I thought they could do better with and offered to specify that we have a bioanalyzer and use it heavily, etc if they thought I was being unfair.

--
Phillip

[KlausR]

I disagree with Eric. We use our Bioanalyzer heavily. But I am far from certain it is a necessity. In fact, it is a rather cantankerous instrument--easily 10% of lanes have bizarre spikes and bulges that disappear when repeated. The reagents are expensive and in some cases go bad even before their expiration date. Their tech support is incapable or unwilling to detail basic, non-intuitive, issues with the function of the instrument. (For example this and this.) Most of what it does can be emulated with minigels and decent gel documentation software.

To be fair, like I said above, we use ours heavily. But if I were starting again I would seriously consider a Caliper instrument instead. And if I were starting a PGM core -- or any other core where $20K was a sizable percentage of the cost of my start up, I would strongly consider doing without.

--
Phillip

I agree with Philip, we have used our Bioanalyzer for a long time and it does give some value, but many lanes fail and more often than not the molecular markers fail the automated detection.