you can use an online tool PlantTFcat (http://plantgrn.noble.org/PlantTFcat/) for analysis and Categorization of Transcription Factor.

Bioinformatic tools will only suggest, not confirm, that your protein is a transcription factor (TF). You got that already, since you know that it is a CCCH-ZF. To confirm that it is a TF, you need on-bench experiments, not bioinformatics.

First up, think of the definition of a TF. A TF is a protein that directly binds a regulatory sequence of a gene, and influences the expression of that gene (activation or repression, or maintaining a basal transcription). [This is in contrast to a co-regulator (CoReg) that can be recruited to DNA segments indirectly, by virtue of its interaction with a TF]. Therefore, you need to split your strategy into two parts: (1) demonstrate that your favorite protein (let's name it YFP1) binds a DNA sequence directly, and then (2) demonstrate that YFP1 influences transcription of a gene containing its binding sequence in the vicinity.

To demonstrate that YFP1 binds DNA directly, you first need to know what sequence it recognizes. That information you can derive bioinformatically, if ChIP-seq datasets are available for YFP1. If there is not, then I fear it will require you to do SELEX-type experiments to identify what sequence it binds. That itself will be a massive project in its own right, and it will also require that you first purify YFP1 to homogeneity. An alternative to SELEX would be a gel-shift assay (also known as gel mobility shift assay or electrophoretic mobility shift assay/EMSA) followed by eluting the slow-migrating DNA and sequencing it. Remember that this could be very challenging, since you should start with a genomic DNA pool sheared to just a few hundred basepairs. That should be your starting material. People did EMSA (and subsequent DNA Footprinting to identify the exact nucleotide sequence bound by the protein) in older days--I might say, in the Jurassic era! Not exactly, but you know what I mean! ChIP-seq is the way to go now.

Once you know the genomic sequence motif recognized by YFP1, you should reconfirm that YFP1 binds that sequence directly. This you can do by following up EMSA with the purified DNA, by using competitor DNA sequences, and also by using a mutated DNA (where the YFP1 binding sequence motif is mutated/deleted). Also, you can generate biotinylated DNA (wt and mutated) by PCR, immobilize them on streptavidin-coated beads, incubate with your purified YFP1, wash, and do immunoblotting to show that YFP binds the DNA.

Next is to demonstrate that it influences transcription. There are again multiple ways to do this. You can generate a reporter construct that contains the wt and mutated YFP1 binding motifs. You can transfect this construct into cells that express YFP1, and quantify the reporter gene expression. In those cells, you can deplete YFP1 (by gene silencing) and re-express it, and show that the reporter gene transcription is dependent on/regulated by YFP1. Next would be to evaluate the expression of a subset of genes where YFP1 binds in the vicinity, and see that their expression changes when YFP1 is depleted.

I am not familiar with the plant cell culture/tissue system; but in principle, you can make nuclear extract (NE) from a cell type that does not express YFP1. You can take a minimal promoter construct (in case you suspect YFP1 to be an activator; or a strong promoter construct in case you suspect YFP1 to be a repressor), and conduct in vitro transcription (IVT) on this construct with NE in the presence or absence of YFP1. Generating a template construct for IVT should be straight-forward, if you get the required sequences correct. The actual IVT reaction and quantifying the RNA is very straight-forward and easy, as well. The only difficulty could be making plant cell NE (or easy; I am unfamiliar with the plant system).

If you do the above in the order, you will have confirmed that YFP1 is a TF. Good luck.

You could try any sequence based function prediction tool for proteins

metaDBsite is a good one

http://projects.biotec.tu-dresden.de/metadbsite/

you can also try

http://www.cogsys.cs.uni-tuebingen.de/software/TFpredict/ http://www.cogsys.cs.uni-tuebingen.de/software/SABINE/.