What would be the best protocol to clone a gene about 5kb in size? The genome is not sequenced, but the gene itself very similar to orthologous genes of organisms with known sequences.
I assume that we are talking about eukaryotic genomes which are split into introns and exons. Depending on what you want to do with your cloned gene, you need different strategies. The method is based on the assumption that the gene sequence has only very little differences to the known sequences of other organisms.
If you want to express your gene from a vector, you need to find the right processed mRNA which is then processed into cDNA and then cloned. Here two strategies are possible: If you have a gene, which is strongly expressed, you can design a primer primer (based on the orthologous sequences) which is anchored in the poly-A tail and has a gene specific seuqence next to it. With this it should be possible to amplify your gene specifically and isolate it. If the gene is not strongly expressed (the resulting amount of cDNA would be very small, since there is no amplification involved), I would opt for making cDNA from total mRNA (use anchored polyA primers) and then amplify the sequence from the cDNA pool with two primers (again based on the orthologous sequences) to amplify the gene sequence. I would go for TA-cloning here, since you do not know, which restriction enzyme recognition sites are present on your gene.
When you need the complete sequence containing introns and exons, there are two possible strategies. First you can use primers to amplify the gene and subsequently clone it. Then it is possible (and this is probably what I would do) to use a set of primers which bind to the end of the gene (and also the reverse strand) to amplify the sequences at the end and then sequence these PCR products. This gives you the exact sequences at the 5' and 3' end of the gene and allows you to include possible regulatory sites, since you then can design exactly matching primers for this region. Again, I would go for TA cloning.
And finally you could start by sequencing the whole region based on homologous sequence primers first and then with primers designed on the newly analyzed sequences. This is the most exact way, but is also costly and takes some time (since you need to make new primers after each round of sequencing until you are done).