Open BinhongLiu opened 2 years ago
BinhongLiu
commented
2 years ago One more question. I encounter an error when running LOCOM with relative abundance data.
res <- locom(otu.table = shotgun.test, Y = Y, fdr.nominal = 0.1, seed = 1, adjustment = "Sandev", n.cores = 4) Error in otu.table + otu.table[, ref.otu] : non-conformable arrays
Could you help with this problem? Thanks
yijuanhu
commented
2 years ago Hi Hongbin,
You raised a very interesting question. LOCOM does not take relative abundance data in its current form, because we use the count data as weights. However, we think it is also reasonable to use relative abundance data directly and weigh each sample equally. We are modifying the code to be updated in a day or two. Hopefully you would still be interested in trying our program.
Per your error, if you remove a sample with zero count data at all taxa, the error should go away.
Yijuan
On Aug 2, 2022, at 7:03 AM, Hongbin Liu @.**@.>> wrote:
One more question. I encounter an error when running LOCOM with relative abundance data.
res <- locom(otu.table = shotgun.test, Y = Y, #C = C[, 1], fdr.nominal = 0.1, seed = 1, adjustment = "Sandev", n.cores = 4) Error in otu.table + otu.table[, ref.otu] : non-conformable arrays
Could you help with this problem? Thanks
— Reply to this email directly, view it on GitHubhttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fyijuanhu%2FLOCOM%2Fissues%2F3%23issuecomment-1202336900&data=05%7C01%7Cyijuan.hu%40emory.edu%7C2de0acddec904e7e3ed908da74769d49%7Ce004fb9cb0a4424fbcd0322606d5df38%7C0%7C0%7C637950350762560815%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=kCy9g%2BAlXWvzuIYhE4wYeV0q6D3rmIdeIw340QDuINw%3D&reserved=0, or unsubscribehttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAFM4UJKI5XWCU4VHZOYOTH3VXD57TANCNFSM55KJDTUA&data=05%7C01%7Cyijuan.hu%40emory.edu%7C2de0acddec904e7e3ed908da74769d49%7Ce004fb9cb0a4424fbcd0322606d5df38%7C0%7C0%7C637950350762560815%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=%2FlTqsgRgt3bOAnu60hNXtbO3q3nUeu1Lysl1VXEZFh8%3D&reserved=0. You are receiving this because you are subscribed to this thread.Message ID: @.***>
BinhongLiu
commented
2 years ago Great! The error was gone. Thanks!
yijuanhu
commented
2 years ago Hi Hongbin,
We investigated the problem of using relative abundance data directly in LOCOM and found it to be nontrivial. For the best performance of LOCOM, we recommend you to look for the original count data, which should exist. Is it possible for you to do that?
Best,
Yijuan
On Aug 3, 2022, at 9:08 AM, Hongbin Liu @.**@.>> wrote:
Great! The error was gone. Thanks!
— Reply to this email directly, view it on GitHubhttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fyijuanhu%2FLOCOM%2Fissues%2F3%23issuecomment-1203927900&data=05%7C01%7Cyijuan.hu%40emory.edu%7C5704624c00ef4de835e808da75513bfe%7Ce004fb9cb0a4424fbcd0322606d5df38%7C0%7C0%7C637951289019962323%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=2F8HaycxBdfhttXLgF0QvY5KxWzBrCH6%2FZ7f7L60NOY%3D&reserved=0, or unsubscribehttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAFM4UJKCNKGV2RK6LMHNTVTVXJVMFANCNFSM55KJDTUA&data=05%7C01%7Cyijuan.hu%40emory.edu%7C5704624c00ef4de835e808da75513bfe%7Ce004fb9cb0a4424fbcd0322606d5df38%7C0%7C0%7C637951289019962323%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=7bXsxy87VyRlGUJpKDyQd4aavb8KZUikzZOJuniFr5U%3D&reserved=0. You are receiving this because you commented.Message ID: @.***>
BinhongLiu
commented
2 years ago The count data is ok for 16s asv table. But for the shotgun metagenome data，it's only a relative abundance table. Was LOCOM applicable for the relative abundance table from shotgun metagenome？
yijuanhu
commented
2 years ago The current version of LOCOM is not. Modifications need to be made for relative abundance data (e.g., either from 16S or shotgun metagenome sequencing).
Yijuan
On Aug 3, 2022, at 12:18 PM, Hongbin Liu @.**@.>> wrote:
The count data is ok for 16s asv table. But for the shotgun metagenome data，it's only a relative abundance table. Was LOCOM applicable for the relative abundance table from shotgun metagenome？
— Reply to this email directly, view it on GitHubhttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fyijuanhu%2FLOCOM%2Fissues%2F3%23issuecomment-1204179172&data=05%7C01%7Cyijuan.hu%40emory.edu%7C06317c004275482249f708da756bbca6%7Ce004fb9cb0a4424fbcd0322606d5df38%7C0%7C0%7C637951402843201063%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=kF%2FqUyNgoQUB%2FJRfrgFvVjBRWi7z8BACeyG7Yxed0Yo%3D&reserved=0, or unsubscribehttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAFM4UJOM6QMZJY6YVO5GPATVXKLTRANCNFSM55KJDTUA&data=05%7C01%7Cyijuan.hu%40emory.edu%7C06317c004275482249f708da756bbca6%7Ce004fb9cb0a4424fbcd0322606d5df38%7C0%7C0%7C637951402843201063%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=kQCtWf%2FokNilVqOf9b6Wx71b4TA4PzbHNcEP2FzrHKo%3D&reserved=0. You are receiving this because you commented.Message ID: @.***>
XbiomeAnalysis
commented
2 years ago Hi，is it possible to add this model mentioned by the following paper to convert relative abundance matrix into counts matrix in LOCOM package?
Paper: Cross-cohort gut microbiome associations with immune checkpoint inhibitor response in advanced melanoma.
Looking forward to your kind reply.
BinhongLiu
commented
2 years ago Hi，is it possible to add this model mentioned by the following paper to convert relative abundance matrix into counts matrix in LOCOM package?
Paper: Cross-cohort gut microbiome associations with immune checkpoint inhibitor response in advanced melanoma.
Looking forward to your kind reply.
Interesting! I'll check this method.
XbiomeAnalysis
commented
2 years ago Hi, @BinhongLiu
Have you achieved the method (logistic-normal distribution modeling)?
Hi, I just found that the count number was used in the example data. Was the LOCOM appliable for relative abundance data? Thanks!