Skip to Content


I just ran an IP and have some bands I want to identify. What do I do now?

Please read the rest of the FAQ. Then call Dr. Hawke at 713-834-6096 to set up an appointment to discuss your project in detail including administrative (IDT and Sample forms to be filled out) and experimental issues.

Is there anything special I should do when running my gel?

Yes! First, the running buffers should be made up fresh for this experiment. Most labs use commercial pre-cast gels, and usually they are OK as long as they have not passed their expiration date. For more tips please read the items below…

How much sample do I need to get a good identification?

A good rule-of-thumb is to have a visible band using coomassie or silver stain, or at least 10-100 fmol of protein. In some cases our usual protease (trypsin) will not be suitable for your protein; this is more likely for small proteins. In such cases more protein or perhaps a different enzyme will be required.

My sample gave a nice, single band by western. Can it be sequenced?

See the answer to Q1 above. If you cannot see it using a typical gel stain, there probably is not enough protein there to identify, or even be sure we are cutting the correct band. Also, a WB only sees your protein of interest, so chances are there are lots of interfering proteins present as well; you simply don’t see them by WB. Further, if you ran whole cell lysate on a gel, blotted it, then ran a Western, there is very little chance we would be able to find anything interesting from that blot, even if you used a non-protein blocking agent. The only realistic chance is to do an IP.

My samples seem to be all keratins. What can I do?

Keratins usually come from human contamination, and can be introduced at almost any point in the handling process. Dust inside buildings is usually mostly human skin flakes, so being careful to avoid dust will help. Keeping samples covered, careful cleaning of all glassware and plastic ware that touches the sample, wearing powder-free gloves, not touching the sample (or sample solutions) with bare hands, avoiding community reagents, and using dedicated equipment can all help to reduce keratin contamination.

What steps does the Facility take to control keratins?

We monitor keratin levels in ongoing sample analyses very closely, and immediately interrupt the sample queue if we suspect a problem. The reagents we use are prepared in large batches, aliquoted, frozen, and carefully tested for performance and contamination. Most of our operations are performed in a standard class-2 hood to reduce contamination. We still normally can detect low levels of keratins in samples we prepare ourselves and process, but it is very difficult to get keratins down to zero!

My samples contain beta-lactoglobulin and other milk proteins. What can I do?

Your lab probably does a lot of western blots using non-fat milk, so mush of the glass- and plastic-ware are probably contaminated with milk proteins. Although this may help you with your WBs, it makes it difficult to do mass spectrometry because the mass spectrometer will “see” peptides from all of these proteins. It is difficult to get glassware clean enough after it has been heavily saturated with protein, a simple solution is to acquire a few key pieces of equipment and reserve them for this purpose ONLY. Then NEVER expose them to milk protein!!

My samples have too much antibody from the immunoprecipitation. How can I get rid of it?

There are a couple of possible strategies to reduce the amount of antibody in the IP elution. One is to use competitive elution if possible, for example using antigenic peptide (FLAG, Myc, are examples). Another is to use covalently bound antibody. If you are working with a construct with a His-tag, the last step could be metal column instead of an antibody.

How can I optimize my IP to get better identification results?

You can perform trial IPs and monitor them by WB. One goal is to pull down as much of the target protein as you can, this you can assess by performing WB on the supernatant. Then, elute as much as you can from the beads, again as determined by WB, multiple elutions from the beads may be performed with the final one using hot SDS-PAGE loading buffer. Finally scale up and try a silver or colloidal coomassie stained gel. Hopefully you also have a representative control for comparison to detect specific binders (assuming you are studying a complex). This gel should give you some idea of the amount of target, as well as “other” wanted or unwanted proteins.

How many cells do I need to grow/harvest to get enough protein for mass spec?

It really depends on how abundant your proteins of interest are. In general you need to use far more lysates for mass spec than you are used to for western blotting. A reasonable starting point may be at least 10x the amount you would use for a western, but further scale-up may be required to see interesting proteins.

If I see a nice band on my gel I can expect to see just one protein from the ID, right?

Wrong!!! We almost always are able to identify several proteins in each band cut from gel in an IP experiment.

What kind of “output” can I expect to get from the Proteomics Facility?

Protein identifications will usually be transmitted as output HTML documents produced by Mascot, our routine search engine. We are happy to discuss these results in detail with you after you have had a chance to review them. Online help for Mascot can also be found on the Matrix Science Web site.

If I see an interesting protein on the list, that means it is really present in my sample, right?

Well, no. Actually it means the protein might be present, the proteins “higher” on the list are more likely to be present, and the ones lower are less likely, depending mainly on the total scores and the scores of the peptides matched. Proteins with scores above about 200 are almost certainly “real”, but as the scores drop, the likelihood of false positives increases. We normally review the weaker hits of interest for you to try to stratify the risk of false positives. A weak match might be weak because there is very little protein present, or it might be a false positive. On request we will give our best estimate of such matches. However, the decision of whether or not to pursue weak matches rests with the investigators.

© 2014 The University of Texas MD Anderson Cancer Center