4.0 KiB
| title | chunk | source | category | tags | date_saved | instance |
|---|---|---|---|---|---|---|
| ChIP-on-chip | 2/3 | https://en.wikipedia.org/wiki/ChIP-on-chip | reference | science, encyclopedia | 2026-05-05T14:02:11.302460+00:00 | kb-cron |
In the first step, the protein of interest (POI) is cross-linked with the DNA site it binds to in an in vitro environment. Usually this is done by a gentle formaldehyde fixation that is reversible with heat. Then, the cells are lysed and the DNA is sheared by sonication or using micrococcal nuclease. This results in double-stranded chunks of DNA fragments, normally 1 kb or less in length. Those that were cross-linked to the POI form a POI-DNA complex. In the next step, only these complexes are filtered out of the set of DNA fragments, using an antibody specific to the POI. The antibodies may be attached to a solid surface, may have a magnetic bead, or some other physical property that allows separation of cross-linked complexes and unbound fragments. This procedure is essentially an immunoprecipitation (IP) of the protein. This can be done either by using a tagged protein with an antibody against the tag (ex. FLAG, HA, c-myc) or with an antibody to the native protein. The cross-linking of POI-DNA complexes is reversed (usually by heating) and the DNA strands are purified. For the rest of the workflow, the POI is no longer necessary. After an amplification and denaturation step, the single-stranded DNA fragments are labeled with a fluorescent tag such as Cy5 or Alexa 647. Finally, the fragments are poured over the surface of the DNA microarray, which is spotted with short, single-stranded sequences that cover the genomic portion of interest. Whenever a labeled fragment "finds" a complementary fragment on the array, they will hybridize and form again a double-stranded DNA fragment.
=== Dry-lab portion of the workflow ===
After a sufficiently large time frame to allow hybridization, the array is illuminated with fluorescent light. Those probes on the array that are hybridized to one of the labeled fragments emit a light signal that is captured by a camera. This image contains all raw data for the remaining part of the workflow. This raw data, encoded as false-color image, needs to be converted to numerical values before the actual analysis can be done. The analysis and information extraction of the raw data often remains the most challenging part for ChIP-on-chip experiments. Problems arise throughout this portion of the workflow, ranging from the initial chip read-out, to suitable methods to subtract background noise, and finally to appropriate algorithms that normalize the data and make it available for subsequent statistical analysis, which then hopefully lead to a better understanding of the biological question that the experiment seeks to address. Furthermore, due to the different array platforms and lack of standardization between them, data storage and exchange is a huge problem. Generally speaking, the data analysis can be divided into three major steps: During the first step, the captured fluorescence signals from the array are normalized, using control signals derived from the same or a second chip. Such control signals tell which probes on the array were hybridized correctly and which bound nonspecifically. In the second step, numerical and statistical tests are applied to control data and IP fraction data to identify POI-enriched regions along the genome. The following three methods are used widely: median percentile rank, single-array error, and sliding-window. These methods generally differ in how low-intensity signals are handled, how much background noise is accepted, and which trait for the data is emphasized during the computation. In the recent past, the sliding-window approach seems to be favored and is often described as most powerful. In the third step, these regions are analyzed further. If, for example, the POI was a transcription factor, such regions would represent its binding sites. Subsequent analysis then may want to infer nucleotide motifs and other patterns to allow functional annotation of the genome.