From High Quality To Perfection

Many researchers perform all kinds of statistical ''tricks,'' such as Lowess correction,1-15-1 to obtain valid data. It is our opinion that by using such tricks, valuable information is lost and that one should strive to solve array artifacts by the technique, rather than by statistics. As an example of this, Fig. 2 shows that ''channel balancing,'' a well-estimated chemical balance between the signal strength in both channels on the array, omits the need for high normalization factors and appears to have an enormous impact on data quality. Such a ''technical'' normalization can be achieved by measuring the amount of label incorporation in both channels and adjusting that for the hybridization. Rather than a conventional spectrophotom-eter, a Nanodrop (www.nanodrop.com) is an excellent piece of equipment for such measurements because it is highly reproducible and needs only tiny amounts of materials. A scatter plot of an array performed without channel balancing shows some sort of ''banana'' shape (Fig. 2A) that could be corrected by Lowess smoothing (Fig. 2B). An array performed with channel balancing (Fig. 2C) does not need such correction.

A second example is the improvement of hybridization conditions by going from a hand-made hybridiza tion chamber[16] to an automated hybridization station (Hybstation12; PerkinElmer, USA). It is our experience that this alleviates the need for localization corrections such as border effects seen in the results of manual hybridizations.

To obtain the best hybridization results, it is essential to estimate hybridization stringency (AT). This is measured by the difference between the hybridization temperature (Texp) and the melting temperature (Tm): AT = Tm—Texp. According to the equation by which Tm can be calculated, the four variables that affect melting temperature are: sodium concentration, formamide concentration, oligonu-cleotide length, and GC content.[17] If AT is too large, the chance of cross-hybridization and background signal is increased. If AT is too small, when the hybridization temperature approaches the melting temperature, too much signal is lost. Thus using the right AT during hybridization and washing can enhance reliability. For our homemade oligo arrays, we use a AT of 15°C. Compare it with a AT of approximately 35-40°C used in Southern blotting to detect gene families and 15-20°C to detect single genes.[17] Thus a AT of 15°C should completely abolish cross-hybridization as long as the design of the oligonucleotides is correct. The design of the oligonucleo-tides will keep pace with the quality of sequence data in the human genome databases. One should realize that things such as time, area, and nature of the surface (i.e., glass or nitrocellulose) are not in the ''Tm equation'' and thus will not affect hybridization stringency.

Finally, RNA quality is of crucial importance to the outcome of an array experiment. An RNA sample that has an excellent quality as judged by the ratio of the 28S and 18S ribosomal RNA bands and, an A260/A280 ratio of 2.0 at a pH 7 usually guarantees good-quality microarray data, provided no phenol traces are present in the RNA samples, which severely affect enzymatic reactions that follow.

Getting Started With Dumbbells

Getting Started With Dumbbells

The use of dumbbells gives you a much more comprehensive strengthening effect because the workout engages your stabilizer muscles, in addition to the muscle you may be pin-pointing. Without all of the belts and artificial stabilizers of a machine, you also engage your core muscles, which are your body's natural stabilizers.

Get My Free Ebook


Post a comment