What is the best tool for estimate the counts of genes with very low expression rate ?

Question

[sabaghianamir70]

i looking for a tool which can give me "count number" for very low expression rate genes or pseudogenes like PER3. What is the best tool for estimate the counts of genes with very low expression rate ? or the best aligner or aligner options to get me maximum amount of fragments to recognize with a FeatureCount or some tools like that.

[ATpoint]

Can you give some details, especially why standard approaches like counting reads over a GTF file that contains these genes with e.g. featureCounts does not help here?

[sabaghianamir70]

i edited my post. i think it is more clear now

[ATpoint]

There is nothing special about lowly-expressed genes when it comes to processing of the data. You can use any of the established tools be it hisat2 followed by featureCounts or salmon followed by tximport. I use the latter one as it takes proper care of multimappers and corrects for GC bias.

[sabaghianamir70]

if i want to get maximum amount of reads, which means map every read even the false one, So. first.:what option should i use(salmon options)? and second: i want filter the false reads with FDR and p value .(is it Right strategy?).i asking because i think you are working on this matter.

[ATpoint]

I strongly recommend against tweaking alignment/mapping parameters that are well-established. I would use salmon to quantify your reads against the reference transcriptome and then take whatever quantification it produces. I do not see the benefit in accumulating false-positives.