JohnBiggs153
commented
1 year ago EDIT 1:
Following the steps of dada2 tutorial, it looks like that possibly that 3% of primers might be causing issues on my processing of the data.
raw_reads filtered ASVs_F ASVs_R joined no_chimera final_perc_reads_retained
Amato170-C16 9591 7348 6862 7004 4265 2969 31.0
Amato169-C12 11753 9016 8346 8683 5037 3713 31.6
Amato168-G6 13588 10709 9853 10231 6228 3845 28.3
Amato167-G2 11647 9258 8516 8785 5091 3333 28.6
Amato166-L3 17799 14500 13582 14099 8278 5786 32.5
Amato165-H2 15473 12174 11421 11843 7078 4350 28.1
Amato192-G10 18482 14711 13740 14143 8641 5048 27.3
Amato164-D1 17969 14315 13515 13938 8625 6146 34.2
Amato191-G5 9896 8020 7313 7649 3670 3017 30.5
Amato190-G1 9233 7403 6721 6973 3493 2662 28.8
Amato189-L2 7076 5595 5191 5327 3018 2076 29.3
Amato188-H1 11108 9265 8660 8939 4928 3650 32.9
Amato186-C22 8906 7048 6437 6679 3532 2617 29.4
Amato185-C14 11852 9706 8869 9301 4856 3322 28.0
Amato184-G8 20392 16273 14961 15642 7915 5449 26.7
Amato183-G4 10996 8420 7624 7952 3857 2916 26.5
Amato182-L5 9173 7569 7005 7311 4530 3535 38.5
Amato181-L1 9861 8090 7517 7858 4434 2723 27.6
Amato163-C25 14043 11240 10284 10736 5937 3925 27.9
Amato180-D3 12587 10247 9547 9899 5594 4127 32.8
Amato179-C24 15949 13066 12196 12614 7710 5134 32.2
Amato178-C21 12372 9833 8979 9349 5186 3617 29.2
Amato177-C13 11875 6659 6006 6338 3127 2246 18.9
Amato176-G7 18838 13911 12721 13343 7003 4422 23.5
Amato175-G3 16275 13391 12166 12942 6252 4858 29.8
Amato174-L4 16232 13186 12373 12820 6691 4802 29.6
Amato173-H3 10321 8556 8123 8274 4985 3114 30.2
Amato172-D2 12226 9676 8971 9255 4816 3397 27.8
Amato171-C23 12480 10098 9218 9613 4813 3661 29.3
Amato162-C15 9644 6558 5961 6218 3352 2519 26.1
Amato161-C11 9977 6829 6169 6432 3132 2282 22.9This is my tracking of the read counts as I go through and process my raw reads. In my opinion it looks like that the read counts look great until we join and remove chimeras from the data. When I was processing this data initially my tracking output looked very similar to this, which lead me to find that I still had my primers attached to my reads. I really don't know what might be causing this issue of my read count dropping after merging and after removing chimeras other than the primers might be causing this issue.
Here is the code for my trimAndFilter along with my merging and chimera removal.
# Actual filtering and trimming part
tnf_summary <- filterAndTrim(F_primers, filt_F,
R_primers, filt_R,
#trimming
truncLen = c(220,170),
#filtering
maxEE = c(2,2),
truncQ = 2)
tnf_summary
# Merging sequences
merged_amp <- mergePairs(dadaF = dada_F,
derepF = derep_F,
dadaR = dada_R,
derepR = derep_R)
seqtab_nochim <- removeBimeraDenovo(seqtab, method = "consensus", verbose=TRUE)
[Forward before trimming]
[Reverse before trimming]
[Forward reads after trim]
[Reverse reads after trim]
benjjneb
I'm trying to analyze a set of Paired FastQ files with dada2 in R. In my analysis I was doing before I realized that the raw reads still had their primers attached. So I went through the ITS workflow to remove my primers. But I'm limited to not using cutadapt function currently. So I was searching the dada2 functions and found the removePrimers function and decided to use that function to help me remove my primers.
With my testing I'm running into interesting issues where removePrimers function doesn't remove all my primers in my fastQ files and creates an imbalance of read counts where I can't use filterAndTrim function after using it.
Notes about data:
Code and Outputs I follow the workflow of the ITS analysis to search my raw reads to see if the primers are in there,
amp_f is a variable where I store all the forward reads and amp_R is a variable where I store all the reverse reads.
So from here we can see the original count of primers in my raw reads. After I get this data I'll show 2 separate methods of removing primers to see which one is the best. Both methods don't remove all primers in my data and I don't know what might be causing this issue.
Test 1, using the removePrimers function
Above is my code block and output to show how I'm using the function and my results. I decided to use the removePrimers function in this way. not including the reverse primer in its parameters since that was causing a huge deletion of my raw reads (file size of my raw reads would go from ~2000-1000 kb in size to 5-1 kb in size files) after the function was done. As you can see this function doesn't even remove all of the primers in my raw reads and after doing other tests with the parameters, like changing the size of the mismatch parameter I can't seem to remove all of the primers in my data.
The other issue when I use removePrimers() is that I can't use filterAndTrim() afterwards, the error that it throws up to me says that the number of reads between the edited forward and reverse files don't match up, like the forward reads have a smaller count compared to the reverse reads Error:
Test 2, using filterAndTrim() to remove primers I saw on this issue page that someone else was having issues with removing all of the primers form their data and the solution for them, if I remember correctly, was using filterAndTrim to remove their primers just be using trimLeft to cut off the first 19 bases for the forward reads and 20 bases for the reverse reads. So I tried to implement that into my data since removePrimers() is giving me some issues.
But we can see that it doesn't remove all of the primers from my data either, it even leaves me with more reads with primers in them than removePrimers(). So I'm currently stuck at a wall where the methods I can use to remove my primers are limited to these 2 options, at least these are the 2 options that I know of that are solely in R.
In my research I know that one of the assumptions for using dada2 is that I need to remove all non-biological nucleotides and I was wondering if having ~3% of the reads (gathered from test 2) will break the dada2 algorithms to help me turn this raw data into data I can put into phyloseq to analyze?