It depends upon which family group you are targeting

If it is Trichoderma species according to our studies TEF 1 alpha gene is most important.

If it is Bipolaris species then the GAPDH gene is important apart from ribosomal universal primers and so on.

Dear Nazia Ji,

It is correct, that Genera or family dependent gen specific amplification.

My doubt was about, what would be the best order in multigene gene sequences in the .fasta file (Ex. ITS, LSU, SSU, RPB2, TEF).

According to certain studies the order like that-ITS LSU SSU TEF RBP2

Thank you, Nazia Ji.

Niranjan M

Hi Niranjan,

There can be pros and cons. But I thought to share my experience.

Number of sequences/genes:

The number of gene regions or sequences is depending on the fungal group that you are working on and the objective of your study. Some gene regions are specific to certain groups. Therefore, better use primary blast searchers (ITS or LSU) to identify the fungal group and then follow past literature to find what are the available gene regions to construct phylogenetic analyses.

If you want to understand the higher-level classification (classes, orders, families), I suggest using more stable gene regions (i.e. exons). When it comes to the lower-level classification such as new species, varieties etc., better use genes that include intra and inter-specific information (i.e. introns, specific protein coding gene regions). Furthermore, increasing the number of sequences in a dataset proved to be beneficial to the predictive power of classifiers.

Length of sequence:

Length of the sequence can be varied with different primers. Easiest way to get an idea is download blast results (20-30 sequences), align them and find out from where most of the others have trimmed. If it is ITS, then you can check feature annotations in the description of blast results. You will find data for some sequences regarding the end of 18s, ITS1, 5.8s, ITS2 and the beginning of 28s. Then you can trim 18s and 28s from the ITS sequences.

Combined gene analyses:

I suggest arranging them according to their stability (evolutionary). Better use exons first and then use introns and protein coding genes at the end.

We need to break taxa according to the Class>Order>Family>Genus>Species>etc.

If we consider ITS, LSU, SSU, RPB2, TEF, and β-Tubulin;

My suggestion is to arrange them as SSU, LSU, ITS (better separate 5.8s and combine after SSU), TEF (if no introns), RPB2 and β-Tubulin

Dear Dhanushka Ji,

Thank you very much for the detailed information.

Cleared most of the doubts.

The SSU, LSU, ITS, TEF, RPB2, and β-Tubulin order can modify

based on an available number of sequences?

"My doubt was about, what would be the best order in multigene gene sequences in the .fasta file (Ex. ITS, LSU, SSU, RPB2, TEF)."

The order of genes in the fungal DNA?

It doesn't matter. The reason is that the software that I use to concatenate genes, it does it in alphabetic order. Since you set the model for each partition, the order is unimportant. I don’t understand why people show the gene boundaries in the phylogeny results. I think it is more reasonable to show the length of each one instead.

Number of sequences/gene?

Does the order depend on the type of family we deal?

The quality of the genes (informativeness level, separability) is more important than the number of genes used. If you use several uninformative genes, you may reconstruct a wrong phylogeny. Generally, a small number of very informative genes produces a phylogenetic tree with better quality. Also, low informative genes will difficult the use of some species delimitation criteria, such as GCPSR.

Length of sequence (Raw or trimmed)?

Just trim parts that seem sequencing errors. Keeping all sequences with the same length implies in reduce informative characters.

Hope it helps!