As all biochemical protocols, the 3C cannot provide for estimating the proportion of cells in which two particular DNA sequences interact, and can only be used to analyzing the average interaction pattern for a given cell population. The portion of cells in which the locus under study has a linear configuration cannot be estimated. If in a given cell population in some cells the locus under study has a looped and in the others – a linear configuration, the profiles of 3C signals will be very similar regardless the ratio of loci with looped and linear configuration. In order to get more insights into the quantitative aspects of the 3C protocol, we have determined absolute yield of ligation products in the 3C analysis of the mouse beta-globin gene locus in erythroid cells. We have measured the input quantity of DNA fragments taken in the 3C ligation reaction and the output quantity of ligation products and have found that the yield of ligation products does not exceed 1% for fragments that are 
assumed to be involved in direct spatial interactions. We reproduced the 3C experiments reported by Tolhius et al. using the restriction enzyme HindIII and analyzed the frequency of ligation of the fragment containing the Hbb-b1 promoter with several selected fragments of the locus. In a conventional 3C experiment, PCR signals from different pairs of primers are normalized to the signals obtained with the random ligation mix�Ca control template prepared by the ligation of equimolar amounts of the DNA fragments of interest, which is assumed to contain equal amounts of all ligation products. Using this standard one can only estimate the AbMole Hexyl Chloroformate relative amounts of ligation products as the copy number of ligation products in the random ligation mix is not known. To determine the absolute yield of 3C ligation products, we prepared a standard equimolar mix of synthesized DNA fragments comprising the exact ligation products with known copy numbers of each fragment, as had been proposed by Comet et al.. More precisely, the mix was composed of PCR-amplified DNA fragments covering ligation junctions of interest. A random ligation mix was used as an amplification template, and the primers used were the same as those used for the subsequent analysis of ligation frequencies. The products of amplification were separated by gel electrophoresis and purified. The obtained template was used to prepare a series of standard dilutions with a known copy number of target DNA sequences; these standards were amplified in parallel with the 3C template. The proximity ligation is the key step in all C-methods. In this work, we have demonstrated that the actual ligation frequencies of the DNA fragments assumed to be assembled in a chromatin hub are in the range of tenths of a percent. This low frequency of ligation was observed between fragments containing the promoter of an active beta-globin gene, Hbb-b1, and its distant enhancers. At the same time, the frequency of ligation of “negative controls”, such as the -42 region and Olfr69 gene, to the Hbb-b1 promoter was below 0.1% in most cases. In our experiments, we used two restriction enzymes, either a standard 6-bp cutter or a frequent cutter. Although the characteristic differences in ligation frequencies observed for different pairs of fragments were well reproduced between the experiments with these endonucleases, the overall level of ligation in the experiment that used MboI digestion was appreciably lower. The low yield of the 3C ligation products observed in our experiments may have different explanations. First of all, the technical reasons should be considered. It is obvious that each restriction fragment has two cohesive ends. If both ends of an anchor fragment and the interacting fragment are taken into consideration, the total yield of the ligation products should be increased about four times.