T-RFLP（末端限制性片段長度多態性）該技術在應用的過程中，肯定需要在實驗條件上不斷進行改進，而這種改進的好壞自然而然需要實驗結果的驗證。
V. Grüntzig在2002年做了該工作的一部分，結果認為，在限制性酶切時，很有必要去除其中影響DNA的酶切過程，并且實驗證明了具體的酶切時間。
具有說服力的結果如下：
1、T-RFLP出數據的速度快，不過只是具有半定量性；
2、DNA片斷的分離有兩種常用的方法：凝膠電泳，近來使用的比較廣泛，且在各種實驗手冊中見到。毛細管電泳。
3、由于一般的實驗程序無法滿足毛細管實驗的要求，因此，作者對該實驗程式進行了改進補充。
4、實驗步驟：將原文步驟貼出來，希望對大家有所幫助。
1) Extraction of community DNA from the sample of interest (soil, sediment, reactor material, water sample, etc). The DNA extraction method of user's choice should yield DNA of high quality lacking inhibitory compounds that could interfere with the subsequent PCR reaction.
2) PCR amplification of gene of interest with fluorescently labeled primers. The T-RFLP method can be applied to any gene, therefore the primer selection will depend on the user's interest. However, 16S rDNA is frequently the gene of choice for microbial community analysis (see PCR protocol). Protocols for specific genes of interest are available elsewhere in literature. The amplification can be performed using a fluorescently labeled forward or reverse primer. Both primers can also be labeled with a different dye and used simultaneously in the same reaction. It might be necessary to pool several PCR reactions to obtain enough product for further steps (200-300 ng of DNA recommended per restriction digest). The amplification efficiency of labeled primers tends to be lower than that of unlabeled primers, frequently leading to lower yields.
3) Concentration and cleaning of PCR products. Excess primers, salts and possible nonspecific PCR products are removed by gel purification. The volume of the pooled PCR reactions can be reduced to half to a fifth of the original volume using a Speedvac or ethanol precipitation in order to facilitate the loading of the sample. Gel purification is then performed by separating the PCR products by electrophoresis on agarose gel, excising the band of proper size and recovering the PCR product with a gel purification kit (Qiagen, MOBIO, Promega) following manufacturer's recommendations. If nonspecific PCR products are not detected, regular PCR clean up kits can also be used.
4) Restriction digestion of the PCR products. Once purified, the PCR products are digested with a restriction enzyme (a four base pair cutter is most appropriate as the probability of having a restriction site within the amplicon is high). Various restriction enzymes can be used in single-enzyme reactions in order to determine which one yields the highest number and most even distribution of terminal restriction fragments. For each digestion, 100-150 ng of purified PCR product (assuming a 50% loss during purification) and 10-20 U of restriction enzyme should be used. The incubation period at the enzyme's optimal temperature can vary from 4-12 h to assure complete digestion. Restriction enzymes are inactivated by heating to 65-80oC for 20-25 min.
5)Desalting of restriction digest. In capillary electrophoresis the injection of DNA samples can be achieved by two methods. First, hydrodynamic injection requires pressure difference over the capillary. Alternatively, electrokinetic injection uses a combination of electrophoresis and electroendosmosis to inject the sample. Applied Biosystems PRISM 310 and 3100 Genetic Analyzers (PE Biosystems) use the latter. The presence of ions can interfere with the uptake of DNA using electrokinetic injection because of preferential injection of higher charge-to-mass molecules (e.g. Clˉ ions). Therefore it is essential to desalt the inactivated restriction digest with Microcon columns (Amicon), Qiaquick Nucleotide Removal Kit (Qiagen) or conventional ethanol precipitation. In our case, the restriction products were diluted with water up to 500 μl, before concentration and desalinization on Microcon columns (see figure below).
6) Capillary electrophoresis (CE); loading. A fifth to a third of the desalted restriction digest is generally loaded onto the capillary electrophoresis system using the default settings optimized for sequencing. This can lead to insufficient fluorescent signal, therefore yielding a low number and height of peaks. The injection time and injection voltage can be varied to regulate DNA uptake. The default sample injection time is generally 10 sec, however longer injection times increase the uptake of DNA from dilute sample solutions, therefore yielding a higher fluorescent signal with more peaks detected. Furthermore, injection voltage is directly proportional to the amount of DNA injected and can be modified from the default of 3 kV. Further concentration of the desalted product may also be necessary to stay within the appropriate CE loading volume.